Real-Time Computer Vision for Human Interfaces · EOG (electro-oculography) Figure: EOG...
Transcript of Real-Time Computer Vision for Human Interfaces · EOG (electro-oculography) Figure: EOG...
EURON Summer School 2003
1
Real-Time Computer Vision for Human Interfaces
Yoshio MatsumotoNara Institute of Science and Technology, Japan
EURON Summer School 2003
2
Introduction to Robotics Lab at NAIST
Nara
EURON Summer School 2003
3
video
EURON Summer School 2003
4
Face Measurement Systemand Its Applications
Yoshio MatsumotoNara Institute of Science and Technology, Japan
EURON Summer School 2003
5
Motivation: Face Measurement
••
• Head motion and gaze direction reflect intention and attention of a person.
• For Human-Robot Interaction,
Goal of this research:
Head Motion => GestureGaze Direction => Attention
=>=>=> Intention
natural ways of communication is required.
EURON Summer School 2003
6
• Head should not move, or head pose is measured by magnetic sensor etc.
• Eye rotation is detected using IR refection on corneal• Head mounted device prohibits natural behaviors • Accurate in best condition, however hard to keep it• Binocular systems and external camera system are also available
Gaze Tracking Techniques:Corneal Reflection
EURON Summer School 2003
7
Gaze Tracking Techniques:Corneal Reflection (cont’d)
• Purkinje images appear as small white dots in close proximity to the (dark) pupil
• tracker calibration is achieved by measuring user gazing at properly positioned grid points (usually 5 or 9)
• tracker interpolates POR on perpendicular screen in front of user
Figure: Pupil and Purkinje images as seen by eye tracker’s camera
EURON Summer School 2003
8
Gaze Tracking Techniques:EOG (electro-oculography)
Figure: EOG measurement:• relies on measurement of skin’s
potential differences using electrodes placed around the eye
• most widely used method some 20 years ago (still used today)
• measures eye movements relative to head position
• not generally suitable for POR measurement (unless head is also tracked)
EURON Summer School 2003
9
Gaze Tracking Techniques: Scleral Contact Lens/Search Coil
Figure: Scleral coil:• search coil embedded in
contact lens and electromagnetic field frames
• possibly most precise• similar to electromagnetic
position/orientation trackers used in motion-capture
EURON Summer School 2003
10
Gaze Tracking Techniques: Scleral Contact Lens/Search Coil (cont’d)
Figure: Example of scleral suction ring insertion:
• most intrusive method• insertion of lens requires care• wearing of lens causes
discomfort
• highly accurate, but limited measurement range (~5 )
• measures eye movements relative to head position
• not generally suitable for POR measurement (unless head is also tracked)
EURON Summer School 2003
11
Real-Time Vision for Face Measurement
• One of the important topics in PUI research• Being actively studied at Microsoft, IBM, MIT,
CMU, Univ. of Illinois etc.
• Few of them can measure the quantitative gaze direction• Most of them use monocular camera systems
EURON Summer School 2003
12
Our Approach
• Stereo camera system for 3D measurement• 3D facial model• Feature tracking by normalized correlation• 3D model fitting algorithm based on spring model
• Real-time Face Tracking• Gaze direction, blinking and lip motion are
additionally measured
EURON Summer School 2003
13
1. Algorithm of Face Measurement
EURON Summer School 2003
14
•••
System Configuration
EURON Summer School 2003
15
3D Facial Model• Corners of eyes and mouth are used for tracking• 3D Facial Model = Template Images + 3D coordinates
EURON Summer School 2003
16
3D Model Fitting for Face Tracking
)()(1
0ii
Tii
n
ii ytRxytRxWE −+−+=�
−
=
),,( zyxT),( ϕφθ ,R
ixiyiW
n
EMinimizing
EURON Summer School 2003
17
d
3D Model Fitting for Face Tracking(cont’d)
),( ϕφθ ,R),( zyxT ,
)0,0,0(R)0,0,0(T
),( ϕϕφφθθ ∆+,∆+∆+R),( zzyyxxT ∆+,∆+∆+
EURON Summer School 2003
18
d
3D Model Fitting for Face Tracking(cont’d)
),( ϕφθ ,−−−R
),( ϕφθ ∆,∆∆),( zyx ∆,∆∆
),( zyxT ,−−−
EURON Summer School 2003
19
Result of Face Tracking
EURON Summer School 2003
20
Accuracy of Face Tracking
Angle of Rotation Stage [deg]
Est
imat
ed A
ngle
[deg
]
0 20 40 60 80
0
20
40
60
80
Measurable Range
Pitch :-80 +80[deg]
Accuracy
Approx. 2[deg]
Yaw :-35 +35[deg]
Roll :-35 +35[deg]
EURON Summer School 2003
21
Estimation of Gaze Direction
EURON Summer School 2003
22
•
•
Result of Gaze Measurement
EURON Summer School 2003
23
Result of Gaze Measurement
EURON Summer School 2003
24
Result of Gaze Measurement
EURON Summer School 2003
25
Result of Face Measurement
EURON Summer School 2003
26
Hardware Configuration
Selectable from below combinations depending on requirements and costs
Desktop PC
Notebook PC
NTSC Camera 2NTSC Camera 2 (Multiplexed)
IEEE1394 Camera 2IEEE1394 High-Speed Camera (80Hz)
EURON Summer School 2003
27
• Multiplexed stereo video streams into a single video stream.
• Vertical resolution of each video signal becomes half.
• Can be used for any conventional image processing system.
Field Multiplexing Device
EURON Summer School 2003
28
The object that a person is looking at is recognized by calculating i for all objects.
Attention Recognition
EURON Summer School 2003
29
Attention Recognition
EURON Summer School 2003
30
Attention Recognition
EURON Summer School 2003
31
Spotting gesture recognition based on Continuous DP Matching using head motions (velocity, angular velocity)
Gesture Recognition
EURON Summer School 2003
32
Gesture Recognition
EURON Summer School 2003
33
Gesture Recognition
EURON Summer School 2003
34
Accuracy
Processing Speed
Specs of Developed System
EURON Summer School 2003
35
Software Configuration
EURON Summer School 2003
36
Software ConfigurationCommercial Face Recognition Software
EURON Summer School 2003
37
Software Configuration
Video Capturing Library
EURON Summer School 2003
38
Software Configuration
Application Software
EURON Summer School 2003
39
Potential Application Areas
Since this system is non-contact, passive, andinexpensive, it can be applied to many application areas where conventional systems cannot be used.
• Human Interfaces• Computer Interface e.g. Hands-free mouse• Robot Interface e.g. Eye-contact communication• Safety System e.g. Driver support• Assistive Products for the disabled
• Human Modeling • Cognitive Science (Experiment on visual cognition• Ergonomics e.g. Human-friendly design)
EURON Summer School 2003
40
2. Application to Human Interfaces
EURON Summer School 2003
41
2.1 Computer Interfaces
EURON Summer School 2003
42
Direct Usage of Head Movements:Hands-Free Mouse
EURON Summer School 2003
43
How to Use Eye Movementsfor Computer Interfaces ?
• Eye movement [Glenstrup 95]– Convergence– Rolling– Saccades– Pursuit motion– Nystagmus– Drift and micro-saccades– Physiological nystagmus
• Need to refine raw data: – distinguish Fixations
from Saccades
EURON Summer School 2003
44
How to Use Eye Movementsfor Computer Interfaces ?
Saccade/Fixation detection• Velocity-Threshold
– Saccades >300 deg/sec.– Fixations <100 deg/sec.– Usual threshold 200
deg/sec.
• Dispersion-Threshold
– Threshold set such that visual angle is between 0.5 ° and 1°.
( ) ( )( ) ( )yy
xxD
minmaxminmax
−−=
EURON Summer School 2003
45
• Command based Interface– Obvious application: Selection of objects {Menu
selection, Window scrolling, …) � Pointing.– Midas Touch problem: Eyes not a control device.– Use dwell time to trigger a selection.
• Non-Command Interfaces– The computer monitors user’s actions instead of
waiting for user’s command.– Potential Applications: User Support
� Detect difficulties and provide translation support of difficult words
How to Use Eye Movementsfor Computer Interfaces ?
EURON Summer School 2003
46
Pro-Active Dictionary:How to detect User Difficulties?
Gaze Pattern in
Normal Reading
Gaze Pattern when
Difficulties Encountered
EURON Summer School 2003
47
Pro-Active Dictionary:Implementation
������������ ���������
�����
����
�������
���������������������
���������
�����
����
����������
�����
��
EURON Summer School 2003
48
Pro-Active Dictionary:Demonstration
EURON Summer School 2003
49
2.1 Robot Interfaces
EURON Summer School 2003
50
Intelligent Wheelchair
PC
CCD cameras
Ultrasonic sensors
Laser Scanner
Notebook PC
CCD cameras
EURON Summer School 2003
51
• Gesture– Nodding -> To start– Shaking ->To stop
• Face Direction– To determine the direction to move
• Gaze Direction– To determine if the user is concentrated
Intelligent Wheelchair
How is facial information used ?
EURON Summer School 2003
52
Intelligent Wheelchair
EURON Summer School 2003
53
Intelligent Wheelchair
Various lighting conditions
EURON Summer School 2003
54
Intelligent Wheelchair
EURON Summer School 2003
55
Intelligent Wheelchair
Sensor-based collision avoidance and wall following
EURON Summer School 2003
56
Result of Experiment (Trajectory)
�����
����
!"
EURON Summer School 2003
57
Result (Input and Output Values)
������������#������������$�����%���������
#��&
������������������'�
( )
�"��*���+
!( !) ,( ,) -( -)
.(
/(
0(
,(
(
1,(
10(
1/(
1.(
*���+
0( 0) )(
2���
3�'�
EURON Summer School 2003
58
Intelligent Wheelchair
-(
,(
!(
(
1!(
1,(
1-(
( -
�"��*���+
*���+
/ 4 !, !) !. ,! ,0
2���������5������� &������
������������6���������7������������6���������7
How to detect concentration of the user?
EURON Summer School 2003
59
Intelligent Wheelchair
Poster
Estimation of User’s Attention
EURON Summer School 2003
60
Intelligent WheelchairEstimation of User’s Attention
Posters
Gaze direction
Robot
EURON Summer School 2003
61
Paying attentionto the poster
Not paying attention to the poster
u Concentrated on the position of the poster.
u Distributed on various positions.
Intelligent WheelchairEstimation of User’s Attention
EURON Summer School 2003
62
Intelligent WheelchairEstimation of User’s Attention
EURON Summer School 2003
63
Receptionist Robot ASKA
EURON Summer School 2003
64
Speech Recognitionand Voice Synthesis
Human andFace Recognition
NLP Dialog Model
Gesture Generation and
Autonomous Mobility
Receptionist Robot
Information Retrievalfrom Databases
Receptionist Robot ASKA
Platform for experimentin real world
EURON Summer School 2003
65
ASKA: System Overview
PC for gesture
Camera
Microphone
Speaker
PC for recognition and speech
PC for vision
PC for server
EURON Summer School 2003
66
Hardware Configuration
telecamera
���1��������"���
Base:tmsuk04 (tmsuk Co.,
LTD.)Degree of freedom
Chest 1 [DOF]Arms 14 [DOF] Hands 6 [DOF]
EURON Summer School 2003
67
Software Configuration
EURON Summer School 2003
68
Receptionist Robot ASKA
EURON Summer School 2003
69
Problem to Solve with Vision
Ø
Ø
••
EURON Summer School 2003
70
Software Configuration
1. Human detection and tracking2. Detection of the sign of utterance
EURON Summer School 2003
71
Pilot Study
••••••
•••
••••
Dialog Experiment
EU
RO
N Sum
mer School 2003
72
Information at U
tterance
EURON Summer School 2003
73
Sign of Utterance : Start
The Face and gaze direct to object before the utterance
EURON Summer School 2003
74
������������������
���������
Sign of Utterance : Start
’•
•
EURON Summer School 2003
75
Experiment : Detection of Utterance Sign
EURON Summer School 2003
76
Demonstration : Interaction Based on Utterance Sign
EURON Summer School 2003
77
3.Application to Human Modeling
EURON Summer School 2003
78
Measurement of Infants
EURON Summer School 2003
79
Measurement of Infants
EURON Summer School 2003
80
Measurement of Infants
Measured Results
Time[sec]Face
Dir
ectio
n (H
oriz
onta
l)
Missing Data Measured Data
EURON Summer School 2003
81
Measurement of TV Game Players
EURON Summer School 2003
82
[s]
[s]
Frequency of Blinking
This result coincides with Psychological Report
Measurement of TV Game Players
EURON Summer School 2003
83
System overview
Measurement of Drivers
EURON Summer School 2003
84
Software Configuration
Measurement of Drivers
EURON Summer School 2003
85
Measurement of Drivers
EURON Summer School 2003
86
Measurement of Drivers
EURON Summer School 2003
87
Experiment at night
Measurement of Drivers
EURON Summer School 2003
88
0
5
0
1
00100 50 0
Hor
izon
tal H
ead
Posi
tion[
mm
]
Time [min]
Yaw
Rate of V
ehicle [deg/sec]0 0.1 0.2 0.3
Ave. Speed 20[km/s] Ave. Speed 40[km/s]
Measurement of DriversHorizontal head movements at curves
Not passive head movements due to centrifugal force,but active head movements to cope with centrifugal force.
0 0.1 0.2 0.3
EURON Summer School 2003
89
0 20 40 60 80
Subject A
Ver
tical
Hea
d Po
sitio
n [m
m]
Subject B Subject C
0 20 40 60 0 20 40 60
Vertical head position in long driving
Measurement of Drivers
0
50
100
Time [min] Time [min]Time [min]
•Body posture changed after long driving ?•Body lowered due to the softness of the seat?
EURON Summer School 2003
90
Measured fixation point and yaw rate of vehicle
Measurement of Drivers
EURON Summer School 2003
91
Attention Recognition
Measurement of Drivers
EURON Summer School 2003
92
Attention Recognition
Measurement of Drivers
EURON Summer School 2003
93
Where are you looking when lane changing?
Measurement of Drivers
Time[s]
EURON Summer School 2003
94
Measurement of Patients in PET
To compensate head motion during measurement
EURON Summer School 2003
95
SummaryHuman Interfaces
Robot InteractionComputer Interface
Face Measurement System
Human ModelingPsychology / ErgonomicsCognitive Science
Accuracy
Processing Speed
EURON Summer School 2003
96
Fin