1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van...

23
1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: [email protected]

Transcript of 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van...

Page 1: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

1

Application-specific constraints for multimedia presentation generation

Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman

CWI Amsterdam

email: [email protected]

Page 2: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

2

Talk overview

•Generating multimedia automatically

•Cuypers multimedia generation engine

•Multimedia and constraints–Quantitative constraints

–Qualitative constraints

•Cuypers demo

•Conclusion, future directions

Page 3: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

3

Generating adaptive multimedia

•Content–Large multimedia database

•System profile–PC, PDA, WAP

•Network profile–Modem, Gigabit

•User profile–Language, Interests, Abilities, Preferences

Too costly to author manually

Page 4: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

4

Cuypers multimedia generation engine

Page 5: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

5

Automatic multimedia generation

•Designer does not specify complete presentation……but only specifies requirements

•System automatically finds a solution which meets requirements

•How should the requirements be specified?–Declarative constraints

Page 6: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

6

Traditional use of constraints

•Constraint solving used for problems with:–Many variables–Large domains

•Based on domain reduction paradigm•Quantitative constraints

– Integer domain–Reduction by arithmetic relations

• Greater than (>)• Less than (<)• Equals (=)

Page 7: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

7

Drawbacks of quantitative constraints

•Too many (trivial) solutions that differ by:–1 pixel position, or–1 milliseconds in timing

•Not sufficiently expressivee.g. cannot specify “no overlap” constraint

•Too low levele.g. A.X2 B.X1

Page 8: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

8

Solution: qualitative constraints

•Example “A not overlap B”, “B after C”•Advantages:

–More intuitive–More expressive–Smaller domains

•Qualitative solutions translate automatically to lower level quantitative constraints

•New problem: What if constraints are insoluble?

Page 9: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

9

Solution: Constraint Logic Programming

•Combine Prolog unification and backtracking with constraint solving

•Use Prolog rules to generate constraints•Backtrack when constraints are insoluble

Page 10: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

10

Cuypers generation engine

•Multiple layers:

–Prolog rules to generate

constraints

–Qualitative constraints

translate to quantitative

constraints

–Solution of both constraints

provides sufficient

information for final

presentation

Page 11: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

11

Cuypers demo: scenario

•Client: User is interested in Rembrandt and wants to know about about the “chiaroscuro” technique

•Server: Query database•Server: Generate constraints according to:

–System profile–User profile–Network profile

•Server: Solve constraints / revise constraints

•Server: Generate SMIL presentation•Client: Play presentation

Page 12: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

12

Conclusions

•Quantitative constraintsare insufficient for automatic multimediapresentation generation. Also need

•Qualitative constraintsto allow intuitive and effectivehigh level specification, and

•Backtrackingfor revising specific constraintsthat cause the entire set to fail

Page 13: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

13

Future directions

•Best-first instead of depth-first–Choose “best” among possible solutions–Needs evaluation criteria

•Improve knowledge management–Make design knowledge declarative and explicit

–Preserve metadata in final presentation–Use standardized and reusable profiles

Page 14: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

14

Page 15: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

15

Need to make trade-offs

•Semantics–Convey message

•Aesthetics–Clear / nice layout

•Resources–Screen size, bandwidth

•Dimension may result in conflicting goals

Page 16: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

16

Quantitative Constraints% csp(+Ids, -Boxes)csp([IdA,IdB],[box(IdA,[x1:AX1, …]), box(IdB,[x1:BX1,…])]) :-

% get valuesmaxX(MaxX), maxY(MaxY),height(IdA,HeightA),widtht(IdA,WidthA),…% define domains

[AX1,AX2,BX1,BX2]::[0..MaxX], [AY1,AY2,BY1,BY2]::[0..MaxY],

% set width & heightAX2 – AX1 #= WidthA,AY2 – AY1 #= HeightA,…% constraintsAX2 #< BX1, % A left-of BAY1 #= BY1, % A top-align B,…true.

Page 17: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

17

Multimedia and Constraints

•Constraint Logic Programming–Domain reduction–Backtracking–Unification (matching rules)

•Qualitative Constraints–Non-integer domain–Allen’s 13 temporal interval relations in three dimensions

Page 18: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

18

Qualitative Constraints

•Example:Two images, A,BA left or right of BA not above or

below B

Page 19: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

19

Qualitative Constraints

% csp(+Ids, -Graph)csp([IdA, Idb], [edge(IdA,IdB,x,NoOverlap),…]) :-

% define domainsNoOverlap :: [b,b-,m,m-],Overlap :: [d,d-,s,s-,f,f-,e],

% constraintsedge(IdA,IdB,x,NoOverlap), % B not-overlap Aedge(IdA,IdB,y,Overlap), % B overlap Atrue.

Page 20: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

20

Qualitative Constraints

•Reasoning– Inverse:

edge(A,B,D,Value) <=> inverse(Value,RValue),edge(B,A,D,RValue).

– Equalityedge(A,B,D,V1), edge(A,B,D,V2) => V1 #= V2

– Transitiveedge(A,B,D,VAB), edge(B,C,D,VBC) =>

tr(VAB,VBC,VAC), % rule generation algorithm

edge(A,C,D,VAC).

•Translation rules to quantitative domainedge(A,B,D,b) => node(A,D/2,V2), node(B,D/1,V1)

V1 #< V2.

Page 21: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

21

Problems in generating multimedia

•Text documents are flexible–Add page, scrollbar,

–Template models

–Wrap text around images

•Multimedia documents are less flexible–No pages or scrollbars, no line-breaking or hyphenation

–Not based on text-flow–Feedback needed

•Linear process model does not work for multimedia

Page 22: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

22

Quantitative Constraints

•Example:Two images,

A,BA left-of BA top-align B

Page 23: 1 Application-specific constraints for multimedia presentation generation Joost Geurts, Jacco van Ossenbruggen and Lynda Hardman CWI Amsterdam email: Joost.Geurts@cwi.nl.

23

Cuypers generation engine

•Rhetoric/Semantic–Sequence, Example

•Communicative devices–Bookshelf, Slideshow

•Qualitative Constraints–A before B

•Quantitative Constraints–A.X2 < B.X1

•Presentation–SMIL