Smart Graphics: 4th International Symposium, SG 2004, Banff, Canada, May 23-25, 2004, Proceedings

172 Pages·2004·3.42 MB·English

by  Andreas Butz

Lecture Notes in Computer Science 3031
CommencedPublicationin1973
FoundingandFormerSeriesEditors:
GerhardGoos,JurisHartmanis,andJanvanLeeuwen
EditorialBoard:
TakeoKanade
CarnegieMellonUniversity,Pittsburgh,PA,USA
JosefKittler
UniversityofSurrey,Guildford,UK
JonM.Kleinberg
CornellUniversity,Ithaca,NY,USA
FriedemannMattern
ETHZurich,Switzerland
JohnC.Mitchell
StanfordUniversity,CA,USA
OscarNierstrasz
UniversityofBerne,Switzerland
C.PanduRangan
IndianInstituteofTechnology,Madras,India
BernhardSteffen
DortmundUniversity,Germany
DemetriTerzopoulos
NewYorkUniversity,NY,USA
DougTygar
UniversityofCaliforniaatBerkeley,CA,USA
MosheY.Vardi
RiceUniversity,Houston,TX,USA
3
Berlin
Heidelberg
NewYork
HongKong
London
Milan
Paris
Tokyo
Andreas Butz Antonio Kru¨ger
Patrick Olivier (Eds.)
Smart Graphics
4th International Symposium, SG 2004
Banff, Canada, May 23-25, 2004
Proceedings
1 3
VolumeEditors
AndreasButz
AntonioKru¨ger
DepartmentofComputerScience,SaarlandUniversity
Stuhlsatzenhausweg,Bau36.1
66123Saarbru¨cken,Germany
E-mail:fbutz,[email protected]
PatrickOlivier
LexicleLtd.
InnovationCenter,YorkSciencePark,York,YO105DG,UK
E-mail:[email protected]
LibraryofCongressControlNumber:2004105605
CRSubjectClassification(1998):I.3,I.2.10,I.2,I.4,I.5,H.5,I.7
ISSN0302-9743
ISBN3-540-21977-3Springer-VerlagBerlinHeidelbergNewYork
Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis
concerned,specificallytherightsoftranslation,reprinting,re-useofillustrations,recitation,broadcasting,
reproductiononmicrofilmsorinanyotherway,andstorageindatabanks.Duplicationofthispublication
orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,1965,
initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer-Verlag.Violationsare
liabletoprosecutionundertheGermanCopyrightLaw.
Springer-VerlagisapartofSpringerScience+BusinessMedia
springeronline.com
(cid:13)c Springer-VerlagBerlinHeidelberg2004
PrintedinGermany
Typesetting:Camera-readybyauthor,dataconversionbyBollerMediendesign
Printedonacid-freepaper SPIN:11005919 06/3142 543210
Preface
The InternationalSymposium on Smart Graphics 2004was held on May 23–25,
2004in Banff, Canada.It was the fifth event in a series which originally started
in 2000as a AAAI Spring Symposium. In response to the overwhelmingsuccess
of the 2000 symposium, its organizers decided to turn it into a self-contained
eventin2001.WiththesupportofIBM,thefirsttwoInternationalSymposiaon
Smart Graphics were held at the T.J. Watson Research Center in Hawthorne,
NY in 2001 and 2002.The 2003symposium moved to the European Media Lab
in Heidelberg to underline the international character of the Smart Graphics
enterpriseanditscommunity.The2004symposiumparticularlyemphasizedthe
contribution of arts and design to the interdisciplinary field of Smart Graphics
andwasthereforeheldattheBanffCentreinAlberta,Canada,aninternationally
recognized center of creative excellence.
The core idea behind these symposia is to bring together researchers and
practitioners from the field of computer graphics, artificial intelligence, cogni-
tive psychology and the fine arts. Each of these disciplines contributes to what
we mean by the term “Smart Graphics”: the intelligent process of creating ex-
pressive and esthetic graphical presentations. While artists and designers have
beencreatingcommunicativegraphicsforcenturies,artificialintelligencefocuses
onautomatingthisprocessbymeansofthecomputer.While computergraphics
providesthetoolsforcreatinggraphicalpresentationsinthefirstplace,cognitive
sciences contribute the rules and models of perception necessary for the design
of effective graphics. The exchange of ideas between these four disciplines has
led to many exciting and fruitful discussions and the Smart Graphics symposia
draw their liveliness from a spirit of open minds and the willingness to learn
from and share with other disciplines.
We would like to thank all authors for the effort that went into their sub-
missions, the program committee for their work in selecting and ordering con-
tributions for the final program, the Banff Centre and our local organizers for
providing space and time for hosting the event, and Springer-VerlagHeidelberg
for publishing the proceedings in their Lecture Notes in Computer Science.
March 2004 Andreas Butz
Antonio Kru¨ger
Patrick Olivier
Organization
Organizing Committee
Conference Chairs Andreas Butz (Saarland University, Germany)
Antonio Kru¨ger (Saarland University, Germany)
Patrick Olivier (Lexicle Limited, UK)
Local Organization Sara Diamond (Banff Centre, Canada)
Brian Fisher (University of British Columbia, Canada)
Program Committee
Maneesh Agrawala (Microsoft Research, USA)
Elisabeth Andr´e (University of Augsburg, Germany)
Steven Feiner (Columbia University, New York, USA)
Sidney Fels (University of British Columbia, Canada)
Mark Hansen (University of California, Los Angeles, USA)
Knut Hartmann (University of Magdeburg, Germany)
Takeo Igarashi (University of Tokyo, Japan)
Rainer Malaka (European Media Lab, Germany)
W. Bradford Paley (Digital Image Design, USA)
Bernhard Preim (University of Magdeburg, Germany)
Thomas Rist (DFKI Saarbru¨cken,Germany)
Andrew Salway (University of Surrey, UK)
Stefan Schlechtweg (University of Magdeburg, Germany)
Sha Xinwei (Georgia Institute of Technology, USA)
Massimo Zancanaro (ITC-IRST Trento, Italy)
Michelle Zhou (IBM T.J. Watson Research Center, USA)
Secondary Reviewers
Blaine Bell, Columbia University
Hrvoje Benko, Columbia University
James T. Klosowski, IBM
Sponsoring Institutions
The 4th International Symposium on Smart Graphics was hosted at the Banff
Centre inAlberta,Canada.Organizationalsupportwasgivenby the BanffNew
Media Institute (BNMI).
Table of Contents
Virtual Characters and Environments
Animating 2D Digital Puppets with Limited Autonomy................ 1
Erin Shaw, Catherine LaBore, Yuan-Chun Chiu, and
W. Lewis Johnson
Non-photorealistic 3-D Facial Animation on the PDA Based on Facial
Expression Recognition ............................................ 11
Soo-Mi Choi, Yong-Guk Kim, Don-Soo Lee, Sung-Oh Lee, and
Gwi-Tae Park
ALTERNE: Intelligent Virtual Environments for Virtual Reality Art .... 21
Marc Cavazza, Jean-Luc Lugrin, Simon Hartley, Paolo Libardi,
Matthew J. Barnes, Mikael Le Bras, Marc Le Renard, Louis Bec, and
Alok Nandi
Tangible and Hybrid Interfaces
Tangible Image Query ............................................. 31
Kreˇsimir Matkovi´c, Thomas Psik, Ina Wagner, and
Werner Purgathofer
Implementation of ActiveCube as an Intuitive 3D Computer Interface ... 43
Ryoichi Watanabe, Yuichi Itoh, Michihiro Kawai, Yoshifumi Kitamura,
Fumio Kishino, and Hideo Kikuchi
Stage-Based Augmented Edutainment ............................... 54
Rainer Malaka, Kerstin Schneider, and Ursula Kretschmer
Graphical Interfaces
Retrieving Vector Graphics Using Sketches ........................... 66
Manuel J. Fonseca, Bruno Barroso, Pedro Ribeiro, and
Joaquim A. Jorge
Quality Metrics for 2D Scatterplot Graphics: Automatically Reducing
Visual Clutter .................................................... 77
Enrico Bertini and Giuseppe Santucci
A View on Views ................................................. 90
Ariel Shamir
Floating Labels: Applying Dynamic Potential Fields for Label Layout ... 101
Knut Hartmann, Kamran Ali, and Thomas Strothotte
X Table of Contents
Poster Presentations
Calligraphic Editor for Textile and Tile Pattern Design System ......... 114
Jos´e Mar´ıa Gomis, Francisco Albert, Manuel Contero, and
Ferran Naya
EvaluationofUser-Friendliness ofa CompactInput Device with Simple
Tactile Feedback.................................................. 121
Itsuo Kumazawa
Petri Net Model for Subjective Views in Collaborative Virtual
Environments .................................................... 128
Jianghui Ying and Denis Graˇcanin
Smart Garden: Plant Mail and Chat Environments.................... 135
Daniel Rivera, Isaac Rudomin, and Marissa Diaz
Automatic Collage Using Texture Synthesis .......................... 140
Stephen Ingram and Pravin Bhat
A Method for Smart Graphics in the Web............................ 146
Thorsten D. Mahler, Stefan A. Fiedler, and Michael Weber
A Framework Supporting General Object Interactions for Dynamic
Virtual Worlds ................................................... 154
Pieter Jorissen and Wim Lamotte
The Media Lounge: a Software Platform for Streamed 3D Interactive
Mixed Media ..................................................... 159
Mark Price
Author Index ................................................ 165
Animating 2D Digital Puppets with Limited
Autonomy
Erin Shaw, Catherine LaBore, Yuan-Chun Chiu, and W. Lewis Johnson
Center for AdvancedResearch in Technology for Education
4676 Admiralty Way,Marina del Rey,90292, USA
{shaw, labore, joechiu, johnson}@isi.edu
http://www.isi.edu/isd/carte/index.html
Abstract. Digital puppetsare animated personas that augment online
educational materials with commentary and summaries. Their anima-
tionsaregenerateddynamicallybasedonuser-authoredtext,contextual
hints, and domain goals, allowing the puppet to act with limited au-
tonomy within specificdomains.Inthispaperwedescribethegraphical
realizationandauthoringofa2Ddigitalpuppet.Wepresentabuild-once,
use-forever production path that allows us to quickly create new char-
acter behaviors, and makes the production of 2D personas feasible. We
describe twodigital puppetapplications and explain howtheanimation
capability is supported across domains.
1 Introduction
Animatedcharactersarebecomingincreasinglypopularforuseinconversational
interfaces,andaspresentationandpedagogicalagents.Inthispaperwedescribe
the graphicalrealizationandauthoringofaanimatedcharacter,orDigitalPup-
pet,whichcanactwithlimitedautonomyinits particulardomain.DigitalPup-
petsaugmentonlineeducationalmaterialswithcommentaryandsummaries,and
havebeenusedinboththetutoringandthepresentationdomain.Theworkpre-
sentedhereisbuildsuponourexperiencescreatingpedagogicalanimatedcharac-
ters,orguidebots [19][20][24][28][12][15].Like guidebots,digitalpuppets interact
with learners via a combination of speech and gestures, making it possible to
more accurately model the kinds of dialogs and interactions that occur during
apprenticeship learning and one-on-one tutoring. They personify the interface,
buildinguponpeople’snaturaltendencytointeractsociallywithcomputers[23],
andcanexpressboththoughtsandemotions,inordertoportrayenthusiasmand
empathy.Likerealexperttutors thatattend toboth motivationalandcognitive
factors,animatedcharactershave the potential to increase learnercuriosityand
interest, and to offer help and reassurancewhen they encounter difficulties [16].
Thedigitalpuppetsystem,andinparticular,itsanimationandauthoring,is
motivatedbytwodrawbackstodevelopingagentsinaresearchenvironment:lim-
itedbudgetsandthe prototypedevelopmentcycle.First,wehavefoundlimited-
animation 2D characters particularly useful for some pedagogical applications;
cartoon2Dforyoungerchildrenandrealistic2Dforadultstorytelling.However,
A.Butzetal.(Eds.):SG2004,LNCS3031,pp.1–10,2004.
(cid:1)c Springer-VerlagBerlinHeidelberg2004
2 Erin Shaw et al.
production demands for animated characters frequently exceed limited research
budgets. This is especially true for complex characters, like the puppets, that
are designed to speak and gesture in parallel. To make the production of 2D
personas feasible, we have developed a build-once, use-forever production path
thatallowsustoquicklycreatenewbehaviorsforagivencharacter[set].Second,
we have found that the prototype iteration cycle for agents precludes a broad
focus: Only after completion of the prototype is thought given to re-purposing
the agent.XML hasmitigatedthe problem,buthasn’tremovedthe challengeof
layperson authoring. Digital puppets were designed to have an integrated user-
friendly authoring environment. We wished to separate the content authoring
andthe contextauthoring,anddidnotwantto distractusers withlow-levelan-
imation details. Puppets speak exactly the lines they are authoredto speak but
their gesturesandfacialanimationaregeneratedautomaticallyfromcontextual
hints; thus, the puppets act with limited autonomy.
2 Related Work
Withrespecttoauthoring,animatedcharactersfallintothreecategories,sequen-
tiallyscriptedcharacters,limitedautonomouscharacters,andautonomousplan-
ning agents. Scriptable online characters such as those from Oddcast, Haptek,
Microsoft have become increasingly common in the commercial domain where
theyactassocialinterfacesorguides[4][13][21].Becauseeachutteranceandges-
turemustbespecified,andbecausetheassociatedgraphicsmustbekeptsimple,
theseinterfaceaidsareoflimitedgeneraluse.Moresophisticatedcommercialap-
plications such as those by Conversive and EDrama apply templates and rules
to alleviate repetitiveness and provide a degree of autonomy [9][11]. Whereas
commercial applications of animated characters assume simple scripting and
graphics, research applications explore multimedia presentation and discourse
planning, and the animation of complex behaviors[3][17][10].
Digitalpuppetsarelimitedautonomouscharacters.Theyarescriptedandse-
quenced, much like in AutoTutor 2 [10][22], though the puppet-scripting model
allows for greater flexibility, supporting parallel actions, behavior choices and
behavior probabilities. The main difference, however, is the use of presentation
goals to create autonomous gestures. In this respect our puppet is similar to
Andre and Rist’s PPP Persona, a multipurpose presentation agent [2][3]. How-
ever, puppets do not plan, and thus are not intelligent or autonomous agents.
Thoughthey are less powerfulwithin any particulardomain, it is fairly efficient
to transfer puppets to new domains.
Theworkpresentedhereiscompatiblewithworkonautomatedbehaviorex-
pression,especially the BodyChat andBEAT systems [6][7]. In BodyChat, gaze
behaviors,for example, are fully automated for turn-taking in multi-participant
settings. Though the Digital Puppets’ behavioral domain is not social commu-
nicationper se,someofits behaviorsaresimilarlyautomated,andcouldbe fur-
ther developed using BEAT. Gestures signifying attentiveness to user actions,
deictic gestures indicating on-screen objects such as lesson pages or simulation