Interactive Evolutionary Computation Hideyuki TAKAGI Kyushu University [email protected]...

Interactive Evolutionary 　 Computation

Hideyuki TAKAGI

Kyushu Universitytakagidesignkyushu-uacjp

http wwwdesignkyushu-uacjp~takagi

Part 1 Humanized Computational Intelligence

1 Humanized Computational Intelligence2 What is IEC3 IEC Applications4 IEC research for Reducing IEC User Fatigue 5 IEC for Human Science6 New IEC framework

1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

neuron model (1943)

Hebbian law (1949)

Perceptron (1958) Adaline

(1960)

Fuzzy Logic (1965)

Fraser (1950s)

GAHolland (1960s)

EPFogel (1960s)

ESRechenberg (1960s-70s)

GPKoza (1980s)

NN-driven Fuzzy Reasoning (1988)

Karr et al (1989)

era of NN FS and EC

era of NN+FS+EC

Computational Intelligence in the 20th Century

Cooperation ofComputational Intelligence

EC

KE FS

NN

Powerful cooperative technologies were developed in the end of 20th Century

What Comes Next

System optimization based on human evaluation

Computer support system for creativity psychological and physical satisfaction

EC

KE FS

NN

+computer real human

Knowledge KANSEI

reasoning

expression

acquisition

associative memory

learning

intuition

preference

subjective evaluation

perception

cognition

etc etc

Two DifferentHuman Capabilities

explicitknowledge

experience

preference

computationalintelligence

optimization

auto-design

implicitknowledge

1st Generation

2nd Generation

3rd Generation

Importance of Human Factors withComputational Intelligence

Humanized technology

ROBOTICSCONTROL

DATAMINING

SIGNALPROCESSING

natural environment human environment

acquired knowledge

physical measurement

qualifying the knowledge

human perception

locationspeedobstacleshellip

preferencesubjective evaluationemotionhellip

database miningIF hellip THEN hellipIF hellip THEN hellip

filter design SN ratio

1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

GPKoza (1980s)


Karr et al (1989)

CI to be a competitor of humans

CI as human models

CI for human

one research direction is

Direction of Computational Intelligence

HumanizedComputational

Intelligence

Interactive Evolutionary Computation

Part 2 What is IEC


What is IEC

Target System

evolutionarycomputation subjective evaluation

(1) fewer population size and fewer generation numbers(2) nevertheless satisfactory solutions can be obtained ie simple landscape(3) discrete fitness in n-evaluation levels and(4) relative fitness in general

Features

searching parameter space

fitne

ss


fitne

ss

normal EC search interactive EC search

Searching spaces of IEC tasks are generally simple

because Any searching points that human operators cannotdistinguish are same for human

80s 90 91 92 93 94 95 96 97 98 99 00 total

graphic art amp CG animation 2 3 2 4 5 5 2 2 4 9 4 423-D CG lighting design 1 3 1 5music 1 3 3 1 1 3 5 17editorial design 1 1 2 4industrial design 2 2 1 5 4 2 4 9 29face image generation 1 1 1 2 1 4 5 1 16speech processing amp prosodic control 2 1 2 1 1 7hearing aids fitting 2 7 5 14virtual reality 1 1 2database retrieval 2 1 8 8 1 20knowledge acquisition amp data mining 5 3 3 1 4 16image processing 1 2 3control amp robotics 1 2 3 4 4 14internet 1 2 1 4food industry 1 1 2geophysics 1 2 3art education 2 2writing education 1 3 4games and therapy 1 1 1 3social system 1 1

discrete fitness value input method 5 2 7prediction of fitness values 1 2 1 8 3 1 16interface for dynamic tasks 1 1 3 5acceleration of EC convergence 1 1 3 1 7combination of IEC and non-IEC 1 2 3active intervention 1 3 2 6

total 2 0 5 5 8 11 23 28 22 48 57 43 252

Statistics of IEC Papers

IEC Research Takagi Lab

(1) 3-D CG lighting design support(2) montage image system(3) speech processing(4) image processing(5) hearing-aidcochlear implant fitting(6) virtual reality in robot control(7) media database retrieval(joint1) virtual aquarium(joint2) geoscientific simulation(joint3) 3-D CG modeling education(joint3) fireworks animation design(joint4) mental disease diagnosis

(1) input interface 11 discrete fitness value input method(2) display interface 21 prediction of users evaluation chars 22 display for time-sequential tasks(3) acceleration of EC convergence 31 approximation of EC landscape 32 IDEgravity IDEmoving vector(4) active user intervention to EC search 41 on-line knowledge embedding 42 Visualized IEC(5) IEC user model(6) New IEC frameworks

application-oriented interface research

IEC for Human Science(1) measuring human mind(2) finding unknown knowledge(3) IEC with physiological responces

IEC Research Categories

art educationwriting educationgames and therapysocial system

speech processinghearing aids fittingvirtual realitydatabase retrievaldata miningimage processingcontrol amp roboticsinternetfood industrygeophysics

graphic art amp CG animation3-D CG lighting designmusiceditorial designindustrial designface image generation

discrete fitness value input methodprediction of fitness valuesinterface for dynamic tasksacceleration of EC convergencecombination of IEC and non-IECactive interventionVisualized IEC

Hideyuki Takagi Interactive Evolutionary Computation Fusion of the Capacities of EC Optimization and Human Evaluationrdquo Proceedings of the IEEE vol89 no9 pp1275-1296 (Sept 2001)

Part 3 IEC Applications



Part 5 IEC for Human Science

CONTENTS

1 Artistic ApplicationsCG lighting design CG animation montage music industrial design

2 Engineering Applicationssignal processing robotics VR media DB retrieval data mining intrusion detection

3 Othersgeo-science coffee blending game amp education

System designed by Tatsuo UNEMI

Interactive GP for Graphic Art

Breeding Forms on the Infinite Planeby W Latham (1992)

Cheerful impressionGloomy impression Heroine movie star Wicked movie star

by HAND

by Interactive GA

IGA for 3D CG Lighting DesignBy K Aoki and H Takagi

Result of Subjective and Statistic Tests

1 level of significance 5 level of significance

cheerful impression

IGA-based lighting design looks effective for beginners

heroine impression

villain impression

gloomy impression

Expanding Lighting EnvironmentsThanks to Increasing LED Performanceand Decreasing LED Costs

Lighting environments expandsthanks to Increasing LED performanceand decreasing LED costs

Increasing the of LED lights makes lighting design difficult

IEC-based Lighting DesignIEC-based Lighting Design

Education of Imagination and Creativity for 3D Shape

It is difficult and time consuming for computer beginners (artist non-technical student etc) to acquire 3D modeling skill

Apply IEC to support CG skill and focus on educating imagination and creativity

3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2

IEC changes shape parameters based on users imagination

Nishino et al (SMC1998 SMC1999)

IEC-base 3D Modeling Concept

designer

3D shape image to create

rateeach

shapevery goodgoodnot good

GeneticAlgorithm

create new 3D geometries by inheriting highly rated shapes and simulating natural evolutionary processes like crossover and mutation

display a set of new shapes

Nishino Takagi and Utsumiya (2000)

Subjective Test

RealisticExamination

Creative Examinationferocious green pepper

Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC

Only 48 (8 parameters times6 primitives) parameters are modified by GA

Internet version of IEC 3-D Modeling Education System

car designers

students artists

IEC-based Modeling Interface

CG System for Car Body

Design

CG System

for Education

CG Systemfor Artwork

Creation

specify rates based on subjective preference

explore CG parameters by evolutionary computation

visualize 3D models by explored parameters

network

display generated models

IEC

Objective Fitness Function

fitness value

Evolution of Funny Animated Figuresby Jeffrey Ventrella

CG Fireworks Animation Made by IEC

type and position

of one of 32

foreheads

shape and position of

nose


mouth


chin

eyes and their

separationX X X X

7 bit 7 bit 7 bit 7 bit 7 bit

= 34 X 1010

faces

Montage Imageby C Caldwell and VS Johnston (1991)

target face montage face three days later 10 GA generations

hair

eyebroweyes

nose

mouth

face basement

Eyes look good

Then Id like to fix the eyes

eyehair

nose mouth

eyebrow

outline

by H Takagi and K Kishi (KES1999)

Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo

by J A Biles (1994)GenJam Jazz Solo Generator

Rhythm Generator by D Horowithz (1994)

Sonomorphs Melody Generator by G L Nelson (1993)

Industrial Designby H Furuta

HTML Design

ECfont of letters colors of letters and background and etc

Looks nice

by N Monmarche et al

Changing Colors and Fonts with IEC

Making Layoutwith IEC

Stability 10Power 05etc

Vigor 10 etc

Completed Poster

By T Obata and M Hagiwara

Color Poster Design

Colors and fonts are VIGOR and layout is STABILITY

Colors and fonts are CLEAN and layout is ELEGANT

Q1 Q2R1

R2

EC

Visitors who click the banner

by R Gatarsk

Evolutionary Banner


CONTENTS




Why IEC-based Signal Processing

There are many cases that SP users are not SP experts but need to design SP filters

image filter hearing aid

Solution is auditory-based SP and visual-based SP without any SP knowledge

IEC realizes this approach

Recovering Distorted Speech

Distorted SpeechOriginal Speech

0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators

Interactive GAa1 a2 a4 a6a0 a7a5a3

8 bits

H(z) = Σa z kk = 0

-k7

Watanabe and Takagi (SMC2005)

Experimental Result

-- Method of Successive Categories --

distance scale

Ideal Situation GA filter designwhen ideal speech is given

40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech

poora little poor neutral a little

good good10｛

Sheffersquos method of paired comparison showed that this difference is significant (plt001)

Experimental Result

Sheffersquos Method of Paired Comparisons

-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech

10th generation of recovering speech



10

20

40

A

B

C

vs vsvsvs vsvs40 1020 4010 20 401020operators

combinations

difference is significant (plt001)

How Distorted Speech was Recovered

Formant area was mainly recovered

(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo

speech synthesizeroriginal

speechspeech analysis

spectral information

ECsubjective evaluation

prosodic control

Voice Conversion by Interactive ECby Y Sato et al

Example of Prosodic Modificationby Y Satoh et al

Original Voice Modified by IEC

Text-to-Speech Intelligible voice

Text-to-Speech male-like voice

Intelligible voice

real voice child-like voice

real voice

>

By T Morita S Iba and M Ishizuka

IEC for Agentrsquos Voice Design

EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB

Hearing Compensation is Difficult

cognitive levelperceptual level

sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed

Tuning System based on How User hears

Ohsaki and Takagi (1998-2007)

Evaluation

evaluation IEC Fitting vs audiologist fitting

monosyllable articulation

sound quality

fitting time IEC lt audiologist

One ndash Two Weeks Later (5 subjects)

Six Months Later (4 subjects)


sound quality

APHAB

Visualized IEC Fitting on a PDA

Visualized EC parameters in an n-D EC landscape are mapped on a 2-D space for visualization

IEC Fitting IEC-based hearing aid fitting

Hayashida and Takagi (IECON2000)

Image Enhancement using

Interactive GPby R Poli and S Cagnoini

Image enhancement filter evolves according to how processed images look well

+ =diastole of patient A systole of patient A

(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1

g1g2 g22 2g2 3g1 g1g2 0854702 min(2g2 2g1 g1(g2 g1))

(1)

(2)

(3)

(4)

(4)

Echo-Cardiographic Image

by R Poli and S Cagnoni (1997)

Non-linear Density Transformation by Using of GP

linear density transformation

non-linear density transformation by using GP

exinvert g(ij)

output

output

input

inputg(ij)= G ( f ( i j ) )

generated by GP

ou

tpu

t d

en

sity

input densityo n f(ij)

n

ou

tpu

t d

en

sity


n

GP-Based Image Filter by Using Neighborhood Pixels Information

f(i j) g(i j)

input pixels output pixels

operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in

G=max[abs(f(ij-1))max-041079+sin(min(f(i+1j-1)-003))-f(i+1j+1)]

Experiment of IGP-Based Edge Detection

input image

Laplacian filter 2nd generation

conventional math-based filter IEC-based filter

high-pass filter 6th generation

Experiment of IGP-Based High Pass Filter Design

input image


x-axis pass filter

input image

Prewitt 4th generation


IGP-Based X-component Extraction

6th generationPrewitt

input image

y-axis pass filter


IGP-Based Y-component Extraction

IEC-based Color Filter Design

input image

examples of coloring

Image Enhancement Filter Design by a Medical Doctor

original ultrasonic image of a lymph node

Enhanced Image after 12 IEC generations

Consumer Robots BusinessHas Been Launched

Home robots will come soon

IEC

Parameters

fearful

secure

Human Friendly Trajectory Control of a Robot Arm

by N Kubota K Watanabe and F Kojim

by H H Lund O Miglimo L Pagliarini A Billard and A Ijspeert

yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts

NN Controller for LEGO Jeep-Robot

1 Children want to make a robot avoiding obstacles

2 Children cannot make a program of its controller but can choose better robot moving

3 Lets evolve the robot controller according to the childrens choice

Interactive EC for Virtual Reality

Arm wrestling human vs robot

controller

fuzzy control rules for VR

IECsubjective evaluation for VR

modifying rules

Kamohara Takagi and Takeda (SMC2007)

IEC for Dental Education in VR

How to realize the reality of VR force feedback

VR dental education software

httpjks-folksstanfordedubonesim

httpelianealhadeffblogspotcom200905dental-serious-games-with-3d-touchhtml

httpwww1800dentistcomdentist-love-blogblogdefaultaspxuseridpost=3

Media DB Retrieval amp Media Converter

NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space

My impression is hellip

by H Takagi et al

IEC-based Image DB Retrievalby S-B- Cho et al

2300 questionnaire for oral care goods

16 image words 2300

16 goods attributes

inductive learning tool

imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n

GA creates variation of image-attribute relation trees

Human chooses simpler and better relations of image-attribute relations

Interactive EC for Data Mining

by T Terano Y Ishino et al

Visual Data-miningThrough 2-D Mapping by GP

X = (x1-u) (873 x12) Y = (467 x6) (252 + 810)for example

by Venturini

Intrusion Pattern Generationsfor Intrusion Detection System

IDS

How to evaluate an intrusion Detection system

Ja-Min KOO and Sung-Bae CHO (2004)

virus

illegal attack

computer virus

How to generate several new types of intrusion patterns for evaluation

virus

virus

J Budynek E Bonabeau and B Shargel (2005)

script individualsgene pool

Generating Intrusion Patternsfor Intrusion Detection Systems

crossover

mutation

deletion




IntrusionDetectionSystem


generatedintrusion patterns

+

GA

detection performancecode evaluation

(A)send ldquolynx zezeonelove64comtemp6htmrrdquoexpect ldquoDownloadrdquosend ldquodrrdquosend ldquo010010010crrdquosleep 1send ldquoyrdquosend ldquoqyrdquo

(B)send ldquovi tempcrrdquosend ldquo1$ s(064)agnrdquosend ldquo1$ s(017)bgnrdquosend ldquo1$ s(003)cgnrdquosend ldquo1$ s(041)dgnrdquosend ldquo1$ s(081)egnrdquosend ldquo1$ s(027)fgnrdquosend ldquo1$ s(055)ggnrdquosend ldquo1$ s(060)hgnrdquosend ldquo1$ s(177)ignrdquosend ldquo1$ s(361)jgnrdquosend ldquo1$ s(011)kgnrdquosend ldquo1$ s(452)lgnrdquosend ldquo1$ s(913)mgnrdquosend ldquo1$ s(784)ngnrdquosend ldquoqnrdquosleep 1

(C)send ldquovi Makefilerrdquosend ldquo1$ shelloctempcgrrdquosend ldquoqnrdquosend ldquomakerrdquosend ldquohellorrdquo

(D)send ldquocp etcshadow etcshaddowrrdquosend ldquochmod o+r etcshadowrrdquo

(E)send ldquorm -rf etcshadownrdquosend ldquocp etcshaddow etcshadowrrdquo

73 of generated executable intrusion patterns were different


CONTENTS




a priori geological knowledge subjective evaluation

Visualization of geological history based on L

Moresis forward code observed current condition

estimated initial condition of

geological features

Evolutionary Computation

geoscientist

mod

el

par

ame

ters

Geological ModelingBased on Interactive EC

Difficulty

bull many variablesndash strength depth stress etc

bull no numerical targetndash Numerical similarity may not mean qualitative similarity such as wrong

fault inclination wrong depth extentndash Help of geologists is necessary

bull computational optimisation with human judgementndash Computational optimisation methods to search material parameters are

required besides geologistrsquos judgement

sketch drawn bya geologist

geologicalforwardmodel

EC evaluation based ongeological knowledge and experience

mat

eria

lpa

ram

eter

s

IEC-based Modelling of Extension of the Earthrsquos Crust

Wijns Moresi Boschetti and Takagi (SMC2001)

IEC-based Modelling of Subduction of Oceanic Crust

animations


ECevaluation based on geological

knowledge and experience

mat

eria

lpa

ram

eter

s

Strength of continentaloceanic crust Convergence rate Depth of sedimentary wedge

Mocha Kilimanjaro

Colombia

Target Brended Coffee

IEC

Fitness Value

Mixture Ratio

Evolving Blends of Coffeeby M Hardy

I go to school on foot

Today I took a bus and

went to school

I enjoyed school swimming

Satomi is a girl who loves athletic meet

She commutes by bus every day

When she arrived at her school

She found there was an athletic meet

Simulated Breeding forComposition Support System

by K Kuriyama T Terano and M Numao

Two sequences of four scenes are chosen as parents

Two better sequencesare chosen as parents

Composition and comparison with similar composition

INITIAL DISPLAY EVOLVED SEQUENCES FINAL DISPLAY

Part 4 IEC research for Reducing IEC User Fatigue


Approaches for Reducing IEC User Fatigue

(1) input interface

11 discrete fitness value input method

(2) display interface 21 prediction of users evaluation chars

22 display for time-sequential tasks

(3) acceleration of EC convergence 31 approximation of EC landscape

32 IDEgravity IDEmoving vector

(4) active user intervention to EC search 41 on-line knowledge embedding

42 Visualized IEC

(5) IEC user model

(6) New IEC frameworks




CONTENTS

1 measuring dynamic range of happy-sad scale

2 finding unknown psychophysiological facts

3 extended IEC IEC with physiological responses

Psychotherapy Diagnostics

IEC-based CGLighting Design

mental patient(schizophrenics)

sad happy

sad happy

emotion range of mentally healthy persons

emotion range of the patient

psychotherapistpsychiatrist

Umm stillhis range is narrower

Takagi Takahashi and Aoki (SMC2004)

PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale

DiscussionDiscussionobtain the order scale of happy - sad expression range from psychological scales of happy and sad obtained through a subjective test

NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS




Cochlear Implants FittingP Legrand C Bourgeois-Republique et al

conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)

485 (practitionerrsquos result)

IEC Breaked Common Sense

0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15

Setting all electrodes but E1E7 E9 under the liminary value gives 821

0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15

Adding electrodes E4 and E15 gives 588 More electrodes better audition

5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15

Only 3 electrodes (E1 E7 E9) have 3 a significant C-T rangebut the best 915

Slide made by Prof Pierre Collet

Obtained Knowledge from Analysis of IEC Fitting

Loudness charact tuned by speech (IEC fitting)ne

fitting using speech1

wwo noise


wwo noise

fitting using speechN

wwo noise

fitting using music wwo

noisene

differences are small

(1)

(2)

The best hearing characteristics of human seems to depend on sounds

EC

hearing aid

Loudness charact tuned by noise or pure tones (conventional fitting)

M Ohsaki and H Takagi (1999)


CONTENTS




EmotionEnhancement

Filter

Example Application

physiologicalresponses

control

physicalfeatures

Emotion Controlwith 3 Steps

IEC with PhysiologyIEC with PhysiologyH takagi (2004)

Step 3 Conventional IEC Based on Psychological Feedback


EC

imagefilters

Conventionalsubjective evaluation

IEC = system optimization technique based on human subjective feedback

Step 3 Extended IEC Based onPhysiological Feedback


EC

imagefilters

Proposal－

physiologicalmeasurement

given targetphysiological

conditions

Part 6 New IEC framework



CONTENTS

1 IEC + EMO

2 Interactive PSO and interactive DE

3 paired comparison-based IEC

Multi-Objective Optimizationwith Subjective Object(s)

Im looking for an apartment that is(1) wider (2) less expensive and (3) closer to a station

wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st

Then how to find an apartment (4) with great view (5) in a quite area

Multi-Objective Optimization MEMS Design

joint research with UC Berkeley

MEMS (Micro Electronic Mechanical Systems) for Sensors Robotics Communications Biotechnology Energy Generation

bull Multi-objective optimization for given specification receiving frequency strength and othersbull We want to use domain knowledge for circuit

design

Raffi Kamalian et al (2004)

bull EMO four objects1 Resonant freq = 100kHz2 K_vertical = 1 Nm3 K_lateral = 100 Nm4 Area = minimized

bull IEC embedding domain knowledge

for MEMS circuit design

IEC + EMO for MEMS DesignIEC + EMO for MEMS Design

IEC+EMO EMOUser Expert of 5s of 5s sign

1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1

Subjective Test + Statistic Testfor EMO+IEC vs EMO


EMO+IEC gt EMO (99 significant by Wilcoxon matched-pairs signed-ranks test )

IEC + EMO for Architectural DesignIEC + EMO for Architectural Design

httpwwwmlitgojpkishakisha0207071220_html

106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Algorithm for Generating LayoutAlgorithm for Generating Layout

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107

IEC +EMO for Room Layout-- Algorithm for Generating Layout --


Makoto Inoue et al (2008)

wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC

IEC +EMO for Room LayoutIEC +EMO for Room Layout

1 room sizes2 room shapes3 moving line to any rooms without

passing through other private rooms4 window sizes requested by

architectural laws5 total water pile length6 total wall length


CONTENTS

1 IEC + EMO



1

32

Particle Swarm OptimizationParticle Swarm Optimization

GA PSOF1 F2 F3 F4 F5 F6 ≒F7 F8

Comparison of Performance

GA vs POS and IGA vs IPOSGA vs POS and IGA vs IPOS( faster ≒ equivalent )

com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess

Tolerance of PSO forQuantization Noise in Fitness


fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41

5段階10段階100段階1000段階

0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels

Experimental Result for F1Experimental Result for F1

Interactive GA

F2 ～ F8 has similar tendency

IGA

IPSO

IPSO with proposed methods

generations

fitne

ssAfter Implementing Four Methods for

Increasing Noise ToleranceNakano and Takagi (CEC2009)


CONTENTS

1 IEC + EMO



Pair Comparison Tournament 　 IECPair Comparison Tournament 　 IEC

Ordinary IEC

Paired Comparison-basedInteractive Differential Evolution


indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F

Takagi and Mitsuyasu (patent)Takagi and Palles(CEC2009)

IGA vs Paired Comparison-based IDE

GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION

bull One of IEC research directions is Humanized Computational Intelligence and IEC is one of tools for this direction

bull IEC has high applicability in wide variety of areas

bull IEC research includes research reducing IEC user fatigue and research for human science besides major IEC research IEC applications

bull We also introduced recent new IEC research on extending IEC algorithms and frameworks

Further Information

bull IEC tutorial amp survey paperndash Hideyuki Takagi Interactive Evolutionary Computation

Fusion of the Capacities of EC Optimization and Human Evaluationrdquo Proceedings of the IEEE vol89 no9 pp1275--1296 (Sept 2001)

ndash Its draft paper is downloadable from the below URL

bull IEC slides are downloadable from the below URL

bull Personal Contactndash takagidesignkyushu-uacjpndash httpwwwdesignkyushu-uacjp~takagi

Interactive Evolutionary　Computation



Cooperation of Computational Intelligence

What Comes Next

Two Different Human Capabilities

Slide 7

Slide 8


Part 2 What is IEC

What is IEC

Searching spaces of IEC tasks are generally simple because





Part 5 IEC for Human Science CONTENTS


Breeding Forms on the Infinite Plane

IGA for 3D CG Lighting Design


Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27


Evolution of Funny Animated Figures


Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner

Part 5 IEC for Human Science CONTENTS (2)



Experimental Result

Experimental Result (2)


Voice Conversion by Interactive EC

Example of Prosodic Modification



Ideal Goal is Far


Evaluation

Slide 51

Image Enhancement using Interactive GP






Slide 58

Slide 59


Slide 61

Consumer Robots Business Has Been Launched



Slide 65




IEC-based Image DB Retrieval


Visual Data-mining Through 2-D Mapping by GP

Intrusion Pattern Generations for Intrusion Detection System

Slide 73

Slide 74

Intrusion Pattern Generations for Intrusion Detection System (2)


Geological Modeling Based on Interactive EC

Difficulty



Evolving Blends of Coffee

Simulated Breeding for Composition Support System






Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98


Part 6 New IEC framework CONTENTS

Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108

Part 6 New IEC framework CONTENTS (2)

Slide 110

Slide 111

Slide 112

Slide 113

After Implementing Four Methods for Increasing Noise Tolerance


Slide 116

Slide 117


CONCLUSION

Further Information



1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

neuron model (1943)

Hebbian law (1949)


(1960)

Fuzzy Logic (1965)

Fraser (1950s)

GAHolland (1960s)

EPFogel (1960s)


GPKoza (1980s)


Karr et al (1989)

era of NN FS and EC

era of NN+FS+EC



EC

KE FS

NN


What Comes Next



EC

KE FS

NN


Knowledge KANSEI

reasoning

expression

acquisition

associative memory

learning

intuition

preference


perception

cognition

etc etc


explicitknowledge

experience

preference


optimization

auto-design

implicitknowledge

1st Generation

2nd Generation

3rd Generation



ROBOTICSCONTROL

DATAMINING

SIGNALPROCESSING


acquired knowledge



human perception





1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

GPKoza (1980s)


Karr et al (1989)


CI as human models

CI for human




Intelligence


Part 2 What is IEC


What is IEC

Target System



Features


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


EC

KE FS

NN


What Comes Next



EC

KE FS

NN


Knowledge KANSEI

reasoning

expression

acquisition

associative memory

learning

intuition

preference


perception

cognition

etc etc


explicitknowledge

experience

preference


optimization

auto-design

implicitknowledge

1st Generation

2nd Generation

3rd Generation



ROBOTICSCONTROL

DATAMINING

SIGNALPROCESSING


acquired knowledge



human perception





1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

GPKoza (1980s)


Karr et al (1989)


CI as human models

CI for human




Intelligence


Part 2 What is IEC


What is IEC

Target System



Features


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

What Comes Next



EC

KE FS

NN


Knowledge KANSEI

reasoning

expression

acquisition

associative memory

learning

intuition

preference


perception

cognition

etc etc


explicitknowledge

experience

preference


optimization

auto-design

implicitknowledge

1st Generation

2nd Generation

3rd Generation



ROBOTICSCONTROL

DATAMINING

SIGNALPROCESSING


acquired knowledge



human perception





1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

GPKoza (1980s)


Karr et al (1989)


CI as human models

CI for human




Intelligence


Part 2 What is IEC


What is IEC

Target System



Features


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Knowledge KANSEI

reasoning

expression

acquisition

associative memory

learning

intuition

preference


perception

cognition

etc etc


explicitknowledge

experience

preference


optimization

auto-design

implicitknowledge

1st Generation

2nd Generation

3rd Generation



ROBOTICSCONTROL

DATAMINING

SIGNALPROCESSING


acquired knowledge



human perception





1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

GPKoza (1980s)


Karr et al (1989)


CI as human models

CI for human




Intelligence


Part 2 What is IEC


What is IEC

Target System



Features


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

explicitknowledge

experience

preference


optimization

auto-design

implicitknowledge

1st Generation

2nd Generation

3rd Generation



ROBOTICSCONTROL

DATAMINING

SIGNALPROCESSING


acquired knowledge



human perception





1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

GPKoza (1980s)


Karr et al (1989)


CI as human models

CI for human




Intelligence


Part 2 What is IEC


What is IEC

Target System



Features


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


ROBOTICSCONTROL

DATAMINING

SIGNALPROCESSING


acquired knowledge



human perception





1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

GPKoza (1980s)


Karr et al (1989)


CI as human models

CI for human




Intelligence


Part 2 What is IEC


What is IEC

Target System



Features


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

1980 1990 2000

NN

FS

EC

NN+FS

GA+FS

NN+FS+GA

GPKoza (1980s)


Karr et al (1989)


CI as human models

CI for human




Intelligence


Part 2 What is IEC


What is IEC

Target System



Features


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Part 2 What is IEC


What is IEC

Target System



Features


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

What is IEC

Target System



Features


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


fitne

ss


fitne

ss




80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

80s 90 91 92 93 94 95 96 97 98 99 00 total



total 2 0 5 5 8 11 23 28 22 48 57 43 252

















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information
















CONTENTS








by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information





by HAND

by Interactive GA




cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



cheerful impression


heroine impression

villain impression

gloomy impression








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information








3D Model Example

blenddeform

P2

P3

P1

P5

P4

P5 P4 P3 P6

P1 P2




designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


designer


rateeach


GeneticAlgorithm




Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Subjective Test



Quality

Operability

Manual gtgt IEC

Manual gt IEC

Manual = IEC

Manual lt IEC



car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


car designers

students artists



Design

CG System

for Education


Creation




network


IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

IEC


fitness value



type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


type and position

of one of 32

foreheads


nose


mouth


chin

eyes and their

separationX X X X


= 34 X 1010

faces



hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

hair

eyebroweyes

nose

mouth

face basement

Eyes look good


eyehair

nose mouth

eyebrow

outline


Montage Image

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

IGA for Music

GenJam

piano drum

code progression

phrase population

measure population

MIDI sequence

bass

goodbad

solo





HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


HTML Design


Looks nice





Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




Vigor 10 etc

Completed Poster


Color Poster Design



Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Q1 Q2R1

R2

EC


by R Gatarsk

Evolutionary Banner


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


CONTENTS











0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information








0

20

40

60

80

100

120

0 1 2 3 4 5 6 7 8 9 10

Pow

er (

dB)

Frequency (kHz)

Distorted Speech

A B

C

three operators


8 bits

H(z) = Σa z kk = 0

-k7


Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Experimental Result


distance scale


40

-20 -10 00 10

A20

C20

A40

C40

B20

B10

B

C10

A10

20

distorted speech


good good10｛


Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Experimental Result


-10 -05 00 05 10

A

B

C 10

20 40 10

20

401020

40

poorer better

distorted speech




10

20

40

A

B

C


combinations




(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



(i) A

A10

A20

A40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(ii) B

B10

B20

B40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

20

(iii) C

20

C10

C20

C40

0 1 2 3 4 5 6 7 8 9 10Frequency (kHz)

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

amplitude

pitch

tempo





prosodic control






Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information





Intelligible voice


real voice

>



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



EC

eval

uatio

n

Proso

dy co

ntro

l

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

I cannot hear both

Hz100 1K 10K

0

20

40

60

80

100

120

dB



sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


sensory level

preference

final hearing

input sounds

measurable in part

target goal

Ideal Goal is Far

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

hearing aid

Conventional

EC

hearing aid

Proposed



Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Evaluation



sound quality





sound quality

APHAB









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information









(1)

(2)

(3)

0832041 g1g2 2g1 2g2 g1

2g2 g2 min( 0277346 g1 g2max(g1g2 g1))

0554695 g1g2 g12g2 g2 g1 g2

g1g2 g22 2g2 2g1g2g1

2 064861 g2 g1


(1)

(2)

(3)

(4)

(4)






exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




exinvert g(ij)

output

output

input


generated by GP

ou

tpu

t d

en

sity


n

ou

tpu

t d

en

sity


n


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


f(i j) g(i j)


operator G

g(ij) = G(f)

x

y

x

y

GP generates G( )

zoom in



input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


input image





input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



input image


x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

x-axis pass filter

input image





input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


input image

y-axis pass filter




input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


input image







IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information






IEC

Parameters

fearful

secure




yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


yj = Σ wijxi + w0j

n

i=1

infrared sensor 1

infrared sensor 2

mechanical switch 1

mechanical switch 2

mechanical switch 3

mechanical switch 4

1 (offset)

6 c

on

tro

l ou

tpu

ts







controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



controller



modifying rules









NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information








NN

psychological

GA search

active

complexity

space factor

pitch

duration

variance

featurespace

space


by H Takagi et al



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



16 image words 2300

16 goods attributes


imag

e w

ords

vs

goo

d at

trib

utes

re

latio

n tr

ee

GA

relation tree 1

relation tree 2

relation tree 3

relation tree 4

relation tree n







by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



by Venturini


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


IDS



virus

illegal attack

computer virus


virus

virus




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




crossover

mutation

deletion







+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information





+

GA









CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


CONTENTS








geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information





geological features


geoscientist

mod

el

par

ame

ters


Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Difficulty









mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




mat

eria

lpa

ram

eter

s




animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


animations




mat

eria

lpa

ram

eter

s


Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Mocha Kilimanjaro

Colombia


IEC

Fitness Value

Mixture Ratio




went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



went to school















(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




(1) input interface







42 Visualized IEC

(5) IEC user model





CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




CONTENTS







sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




sad happy

sad happy






PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

PM

PK

NH

NS

NY

NK

happy impression

NN

PT

happy scale

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

NY

NK

NN

NS

NH

PM

PK

PT

sad impression

sad scale


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


NS

NN

NK

PM

NY

PT

NH

PK

1st 5th

2nd

3rd

4th

6th

7th

8th

sad happy sad happy


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


CONTENTS





conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


conventionalfitting

E1 E4 E5E2 E3 E6 E7 E9E8 E13E14E15

20

15

10

5

electric channels

inte

nsity

IEC fitting

eval

uatio

nfitting

45

915 (IEC fitting)



0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


0

5

10

15

20

25

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15


0

2

4

6

8

10

12

14

16

18

20

22

1 2 3 4 5 6 7 8 9 13 14 15


5678910111213141516171819202122

E 1 E 2 E 3 E 4 E 5 E 6 E 7 E 8 E 9 E 13 E 14 E 15






wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




wwo noise


wwo noise


wwo noise


noisene


(1)

(2)


EC

hearing aid




CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


CONTENTS




EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

EmotionEnhancement

Filter

Example Application


control

physicalfeatures





EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



EC

imagefilters





EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



EC

imagefilters

Proposal－



conditions




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




CONTENTS

1 IEC + EMO





wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



wid

th

tenant width

dist

ance

to

st

tena

nt

distance to st






design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information





design







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information







1 Y 7 9 -12 Y 12 6 13 Y 7 3 14 N 6 2 15 Y 4 4 06 Y 11 9 17 N 8 7 18 Y 1 0 19 N 6 3 110 N 12 7 111 N 9 2 1






106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



106

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda


Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda

Corridor

W

K

Wall

L

B

Verenda 107




wal

l

wal

l

wal

lw

all

wal

l

wal

l

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Corridor

L B

W

Wall

K

Verenda

Corridor

W

B

L Wall

K

Verenda

Corridor

W Wall

B K

Verenda

Corridor

L W Wall

B K

Verenda

Corridor

L W

B Wall

Verenda

Corridor

W

K

B Wall

L

Verenda

Layout Algorithm

objective 1

obje

ctiv

e 2

EC

EMO IEC






CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


CONTENTS

1 IEC + EMO



1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

1

32





com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information




com

plex

ity

high

low

IGA IPSO

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

54321

5-levels

10-levels

100-levels

1000-levels

quan

tizat

ion

nois

ein

fitn

ess



fewer

more

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

0

1

2

3

4

5

6

7

1 11 21 31 41


0

1

2

3

4

5

6

7

1 11 21 31 41


Interactive PSO

levels

levels

levels

levels

levels

levels

levels

levels


Interactive GA


IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

IGA

IPSO


generations

fitne




CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


CONTENTS

1 IEC + EMO




Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


Ordinary IEC



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information



indi

vidu

als

targetvector

trialvector

targetvector

parametervector 1

parametervector 2

basevector

mutantvector

F



GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information


GA

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

HA 1

HA 2HA 1DE

ON ON

HA 5 HA 6

HA 9 HA 10 HA 11 HA 12

HA 13 HA 14 HA 15 HA 16

HA 3

HA 8

HA 2

HA 7

HA 4

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

CONCLUSION





Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Further Information










What Comes Next


Slide 7

Slide 8


Part 2 What is IEC

What is IEC











Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27




Montage Image

Slide 32

IGA for Music

Industrial Design

HTML Design

Color Poster Design

Evolutionary Banner




Experimental Result







Ideal Goal is Far


Evaluation

Slide 51







Slide 58

Slide 59


Slide 61




Slide 65








Slide 73

Slide 74




Difficulty










Slide 88

Slide 89

Slide 90


Slide 92




Slide 96

Slide 97

Slide 98



Slide 101

Slide 102

Slide 103

Slide 104

Slide 105

Slide 106

Slide 107

Slide 108


Slide 110

Slide 111

Slide 112

Slide 113



Slide 116

Slide 117


CONCLUSION

Further Information

Interactive Evolutionary Computation Hideyuki TAKAGI Kyushu University [email protected]...

Documents

Transcript of Interactive Evolutionary Computation Hideyuki TAKAGI Kyushu University [email protected]...