IVA09 - 9th International Conference on Intelligent Virtual Agents


IVA 2010
IVA 2008
IVA 2007
ACII 2009
GALA 2009
BML Workshop

Latest news

Oct 2, 2009:
Online version of the proceedings at Springer

Sep 22, 2009:
Links to PDFs of accepted papers are being added. Find them on the Programme webpage or in Programme PDF.


9th International Conference on Intelligent Virtual Agents - IVA'09


Intelligent virtual agents (IVAs) are interactive characters that exhibit human-like qualities and communicate with humans or  with each other using natural human modalities such as  speech and gesture. They are capable of real-time perception, cognition and action that allows them to participate in a dynamic social environment.

IVA-09 is an interdisciplinary annual conference and the main forum for presenting research on modeling, developing and evaluating intelligent virtual agents with a focus on commu-nicative abilities and social behavior. In addition to pre-sentations on theoretical issues, the conference encourages the showcasing of working applications. Researchers from the fields of human-human and human-robot interaction are also welcome to share work which has a bearing on intelligent virtual agents.

IVA-09 will be co-located in Amsterdam with the complemen tary Affective Computing & Intelligent Interaction International Conference (ACII-09), held 10-12 Sep 09. While this is a great opportunity to make the most of a trip to Amsterdam, the same paper should not be submitted to both conferences: papers with a primary focus on intelligent virtual agents are encouraged to be submitted to IVA, papers with a  focus on emotion are encouraged to be submitted to ACII.

Download call for papers.

Special topic: Games

IVA-09 particularly encourages submissions on this year's special topic of games. Read more.

In cooperation with

Cooperation partners

Sponsored by



Design and modeling of IVAs

Implementation of IVAs

Applications of IVAs

Conceptual frameworks for IVAs

Special topic: Games

IVA-09 particularly encourages submissions on this year's special topic of games. Read more.

Invited Speakers

Casey Hudson

project director at BioWare

Talk Title: Past and Future Challenges in Creating Emotionally-Engaging Real-Time Digital Actors in Videogames


Evolving beyond their origins as a novel pastime, videogames have developed into a medium with tremendous power to entertain and engage players through emotionally powerful interaction. These emotional connections are often powered by the quality of the digital actors that inhabit game worlds and bring them to life. But as technologies for creating lifelike characters escalate, so do the challenges of the creation process. This discussion examines methods used by cutting-edge games to create deeply compelling digital actors, and explores future challenges and solutions that will help videogames unlock the full potential of emotionally engaging human interaction.


Casey Hudson has been a videogame developer for over 10 years, contributing to some of the industry's most critically and commercially successful titles. Since his start in the business he has worked at BioWare in Edmonton, Canada. BioWare is a world leading studio in the development of computer, console, handheld, mobile and online roleplaying games, building emotionally engaging interactive entertainment focused on rich stories and memorable characters. Since 1995, BioWare has created some of the world's best-selling titles, including the award-winning Baldur's Gate and Neverwinter Nights series, Star Wars: Knights of the Old Republic and Mass Effect. Upcoming BioWare titles include Mass Effect 2, Dragon Age: Origins, and Star Wars: The Old Republic.

Casey started his career in game development as a Technical Artist, creating 3D art and technical solutions for MDK2 (2000) and Neverwinter Nights (2002). Casey was the Project Director on Star Wars: Knights of the Old Republic (2003), winner of over 40 Game of the Year awards - and one of the highest-rated videogames of all-time. As the Project Director on the New York Times' 2007 "Game of the Year" Mass Effect, Casey led a team in the development of an ambitious game set in an all- new science fiction universe, which pushed the boundaries of interactive cinematic storytelling and advanced digital actors. He is currently leading the development of the second title in the Mass Effect trilogy, and manages the Mass Effect franchise as its Executive Producer. Casey Hudson received his BSc. in Mechanical Engineering from the University of Alberta in 1998.

Steve Di Paola

artist, scientist and associate professor (homepage)
at Simon Fraser University, Canada

Talk Title: Intelligent Expression-based Character Agent Systems: An Interdisciplinary Approach


By using parameterization techniques which model artistic, living or cognitive systems it is becoming possible to create new types of character expression systems. These new emerging techniques are allowing virtual agent creators to incorporate models of expression, emotion, behavior and even human creativity into their creations. Rather than simply using realism as a goal, is it possible to computation model knowledge from artists, musicians and zoologists to go from communication to expression based agent systems. Steve DiPaola who works as a computer scientist, researcher and artist will discuss and demonstrate his work and his future vision in these emerging areas (see ivizlab.sfu.ca). DiPaola will demonstrate his lab's voice and behavior driven 3D facial system, and its application in psychology research, gaming and agent systems. He will demonstrate his collaborative work with industry like with the Vancouver Aquarium creating an Artificial Life based Virtual Beluga Whale Interactive, a system that allows onlinegamers to browse an evolving face-space for the EA/Maxis game 'The Sims', as well as more experimental systems that explore computer creativity by extracting emotion out a music score or automatically evolve (Genetic Programming) non photo-realistic characters (dipaola.org/evolve) as a way to incorporate automatic computer creativity.


Steve DiPaola, equally active as an artist and scientist is an Associate Professor at Simon Fraser University in Vancouver. He directs the iViz research lab which strives to make interactive systems bend more to the human experience by incorporating parameterized cognitive and living system models. He came to SFU from Stanford University and before that spent 10 years as a senior researcher at NYIT Computer Graphics Lab, an early pioneering lab in high-end 3D techniques. He has held leadership positions (CTO, VP Creative, Dir of Dev.) at leading edge companies including Electronic Arts (Advanced Technology Group), Saatchi & Saatchi Innovation, Silicon Valley virtual agent start-ups and has consulted for HP, Kodak Research, Macromedia and the Institute for the Future. His still, interactive and performance-based art work has been exhibited international including the AIR and Tibor de Nagy galleries in NYC as well as the Whitney Museum of Art, the Smithsonian, IBM Gallery of Science and Art and more recently at teh MIT Museum. He has created collaborative pieces with media artist Nam June Paik and electronic music pioneers Kraftwerk and is known for making new media authoring tools used equally by artists, scientists and universities. See dipaola.org.

Marilyn Walker

professor (homepage)
at University of California, Santa Cruz, USA

Talk Title: Endowing virtual characters with expressive conversational skills


When humans interact with one another, socially intelligent behaviors arise from the interaction of personality, cultural knowledge, the ability to observe and reason about social relationships, and the ability to project and detect affective cues. Thus one important component for endowing virtual agents with social intelligence is a mechanism that provides parameters to represent both personalities and social relationships, and that makes it easy to vary these parameters in order to affect agents' linguistic behaviour in dialogue with human users. In this talk, I will discuss how our work aims for both psychological plausibility and realistic usability. To achieve psychological plausibility we build on theories and detailed studies of human language use, such as the Big Five theory of personality, and Brown & Levinson's (1987) theory of politeness. To achieve realistic usability, we develop both rule-based and trainable methods that can dynamically, and in real time, change an agent's linguistic style by modifying the values of theoretically motivated parameters, and we show empirically that these variations are perceived by human users as the agent intended.


Professor Marilyn Walker is a computational linguist and computer scientist and holder of a Royal Society Wolfson Research Merit Award at the University of Sheffield. She holds a B.A. with Honors from UC Santa Cruz (1984), an M.S. from Stanford University (1988) and a Ph.D. from University of Pennsylvania (1993). At Sheffield, she leads the Cognitive Systems group, whose research focuses on endowing Virtual Agents with the ability to be socially aware, to engage the user, to be cooperative, and to learn from interactions with the user to improve the system's interactions. She has been conducting research for almost 25 years on dialogue technologies at Hewlett Packard Labs, Mitsubishi Electric Research Labs, AT&T Labs Research, and has written over 100 refereed papers and patents. In 1995, for the VIVA virtual theatre, her group developed the first automatic spoken language generator based on Politeness theory that allows virtual agents to produce language displaying aspects of human politeness behaviour. In 1997, she developed the first application of reinforcement learning to optimize dialogue strategy selection, by interacting with human users in the ELVIS system (EmaiL Voicemail Interactive System). Her group recently developed Personage, a highly flexible and completely trainable language generator that can choose what personality to manifest, for example, whether to be extraverted or introverted. This research has applications in dialogue systems, computer gaming, interactive drama systems, virtual training environments (serious games), and intelligent tutoring systems.

Special topic: Games

IVA-09 particularly encourages submissions on this year's special topic of games. The game industry is the source of the world's largest selection of interactive characters. To date, the creation of these characters and their social behavior has largely relied on carefully hand crafted techniques rather than automation. With larger environments, grander stories, more players and a greater demand for realism, hand crafted approaches are unlikely to scale. Imbuing characters with more intelligence and self-determination is an ongoing and so far unfulfilled goal of the game industry. IVA-09 is an opportunity to reveal, tackle and discuss the issues that relate to using intelligent virtual agents in games and aims to strengthen links and an the exchange of knowledge between academia and the game industry.


NEMOIVA 2009 will be held in Amsterdam, in NEMO, the largest science centre in The Netherlands, where science and technology come to life in interactive exhibitions. NEMO is located near Amsterdam Central Station in the old harbour in an eye-catching green building that is often compared to a ship. The building is a key piece of modern architecture designed by the well known Italian architect Renzo Piano whose work includes the design of the Centre Pompidou in Paris.

Amsterdam is one of the most beautiful cities in Europe. It has a large historical inner city with fascinating architecture and a network of canals spanned by hundreds of bridges. It has many world-famous museums and international restaurants. A tourist guide to Amsterdam can be found here or here.


A limited number of hotel rooms has been reserved for the conference in the CitizenM hotel, locations Amsterdam City and Amsterdam Airport, from September 9th up to and including September 15th (check-out date September 16th). The price of these rooms is 139 Euros. If you want to book one of these rooms please contact Alice Vissers/Charlotte Bijron Alice Vissers/Charlotte Bijron until Juli 12th. Please state the location you prefer (City or Schiphol Airport) and which nights you would need a hotel room.

Due to a very big conference taking place in Amsterdam almost simultaneously with IVA we have not been able to reserve rooms in other big hotels in the center of Amsterdam. Private bookings in other hotels are still possible but there are not many rooms left. We strongly advise you to book as soon as possible. An online hotel booking service can be found here and here.

Please check here whether or not you require a visa for the Netherlands.

Special Event

On Tuesday September 15th, the conference dinner will take place in Artis, the oldest zoo in the Netherlands, located in the heart of Amsterdam.


Best Paper Award

The Best Paper Award received:
  • Engagement vs. Deceit: Virtual Humans with Human Autobiographies - Timothy Bickmore, Daniel Schulman, Langxuan Yin
The following papers were nominated:


IVA GALA Awards can be found here.


Programme in PDF

Online version of the proceedings at Springer

List of accepted papers and posters (PDF)

Schedule and Programme (PDF)

Schedule of IVA09

  14 Monday 15 Tuesday 16 Wednesday
9 Local organizer present Invited talk 2
S. Di Paola
Invited talk 3
C. Hudson
9.30 Registration, Coffee, Welcome
10 Invited talk 1
M. Walker
Coffee break Coffee break
10.30 Paper session 4
Facial Expression;
Paper session 7
Motion capture
11 Paper session 1
Personality and Memory;
Mindful Agents
12 Paper session 8
Dialogue and Speech
12.30 Lunch Lunch
13 Good-bye (in foyer)
Paper session 2
Gesture and
Bodily Behaviour
Paper session 5
Affect and Empathy
15.30 Coffee break Coffee break  
Paper session 3
Paper session 6
Agents in Virtual Worlds and Games
17.30 Poster session
Possibility to show demos;
reception at NEMO
  Dinner, awards at Artis  



Program of paper sessions IVA09


Invited talk 1, 10.00-11.00:
Endowing virtual characters with expressive conversational skills Marilyn Walker

Session 1, 11.00-12.30: Personality and Memory

  1. Engagement vs. Deceit: Virtual Humans with Human Autobiographies
    Timothy Bickmore, Daniel Schulman, Langxuan Yin
  2. A Socially-Aware Memory for Companion Agents
    Mei Yii Lim, Ruth Aylett, Wan Ching Ho, Patricia Vargas, Sibylle Enz
  3. A Model of Personality and Emotional Traits
    Margaret McRorie, Ian Sneddon, Etienne de Sevin, Elisabetta Bevacqua, Catherine Pelachaud
  4. BDI-Based Development of Virtual Characters with a Theory of Mind
    Michal Sindlar, Mehdi Dastani, John-Jules Meyer
  5. How Place and Objects Combine? What-Where Memory for Human-like Agents
    Cyril Brom, Jiri Lukavsky
  6. EXSTASIS - An Extended Status Model for Social Interactions
    Martin Rumpler
  7. Authoring Behavior for Characters in Games Reusing Abstracted Plan Traces
    Antonio A. Sánchez-Ruiz, David Llansó, Marco Gómez-Martín, Pedro A. González-Calero

Session 2, 13.30-15.30: Gesture and Bodily Behaviour

  1. Modeling Peripersonal Action Space for Virtual Humans Using Touch and Proprioception
    Nhung Nguyen, Ipke Wachsmuth
  2. GNetIc -- Using Bayesian Decision Networks for Iconic Gesture Generation
    Kirsten Bergmann, Stefan Kopp
  3. A Probabilistic Model of Motor Resonance for Embodied Gesture Perception
    Amir Sadeghipour, Stefan Kopp
  4. A Groovy Virtual Drumming Agent
    Axel Tidemann, Pinar Öztürk, Yiannis Demiris
  5. Motion Synthesis Using Style-editable Inverse Kinematics
    Gengdai Liu, Zhigeng Pan, Ling Li
  6. Methodologies for the User Evaluation of the Motion of Virtual Humans
    Sander Jansen, Herwin van Welbergen

Session 3, 16.00-17.30: Evaluation

  1. A Study into Preferred Explanations of Virtual Agent Behavior
    Maaike Harbers, Karel Van den Bosch, John-Jules Meyer
  2. Evaluating Adaptive Feedback in an Educational Computer Game
    Cristina Conati
  3. Media Equation revisited. Do users show polite reactions towards an embodied agent?
    Laura Hoffmann, Nicole Kraemer, Anh Lam-Chi, Stefan Kopp
  4. The Lessons Learned in Developing Multi-user Attentive Quiz Agents
    Hung-Hsuan Huang, Takuya Furukawa, Hiroki Ohashi, Aleksandra Cerekovic, Yuji Yamaoka, Igor Pandzic, Yukiko Nakano, Toyoaki Nishida
  5. On-Site Evaluation of the Interactive COHIBIT Museum Exhibit
    Patrick Gebhard, Susanne Karsten
  6. Evaluating an algorithm for the generation of multimodal referring expressions in a virtual world
    Werner Breitfuss, Ielka van der Sluis, Helmut Prendinger,  Saturnino Luz, Mitsuru Ishizuka

Demo and poster session, 17.30-19.00


Invited talk 2, 9.00-10.00:
Intelligent Expression-based Character Agent Systems: An Interdisciplinary Approach Steve Di Paola

Session 4, 10.30-12.30: Facial Expression and Gaze

  1. Expression of Emotions using Wrinkles; Blushing; Sweating and Tears
    Celso de Melo, Jonathan Gratch
  2. Impact of Expressive Wrinkles on Perception of Facial Expressions of Emotions by a Virtual Character
    Matthieu Courgeon, Stéphanie Buisine, Jean-Claude Martin
  3. Real-time Crying Simulation for 3D Characters
    Wijnand van Tol, Arjan Egges
  4. Breaking the Ice in Human-Agent Communication: Eye-Gaze Based Initiation of Contact with an Embodied
    Nikolaus Bee, Elisabeth André, Susanne Tober
  5. An Approach for Creating and Blending Synthetic Facial Expressions of Emotion
    Meeri Mäkäräinen, Tapio Takala
  6. Animating Idle Gaze in Public Places
    Angelo Cafaro, Raffaele Gaito, Hannes Vilhjalmsson

Session 5, 13.30-15.30: Culture, Affect and Empathy

  1. Virtual Agents and 3D Virtual Worlds for Preserving and Simulating Cultures (video)
    Anton Bogdanovych, Juan Antonio Rodriguez,  Simeon Simoff, Alex Cohen
  2. One for all or one for one? The influence of Cultural Dimensions in Virtual Agents' Behaviour
    Samuel Mascarenhas, Joao Dias, Rui Prada, Ana Paiva
  3. Combining Facial and Postural Expressions of Emotions in a Virtual Character
    Céline Clavel, Justine Plessier, Jean-Claude Martin, Laurent Ach, Benoit Morel
  4. Expression of Moral Emotions in Cooperating Agents
    Celso de Melo, Liang Zheng, Jonathan Gratch
  5. Evaluating Emotive Character Animations Created with Procedural Animation Yueh-Hung Lin, Chia-Yang Liu, Hung-Wei Lee, Shwu-Lih Huang, Tsai-Yen Li
  6. Modeling emotional expressions as sequences of behaviors
    Radoslaw Niewiadomski, Sylwia Hyniewska, Catherine Pelachaud
  7. I Feel what you Feel: Empathy and Placebo Mechanisms for Autonomous Virtual Humans
    Julien Saunier, Hazaël Jones, Domitile Lourdeaux
  8. Predicting User Phsychological Characteristics from Interactions with Empathetic Virtual Agents
    Jennifer Robison, Jonathan Rowe, Scott McQuiggan, James Lester
  9. When Human Coders (and Machines) Disagree on the Meaning of Facial Affect in Spontaneous Videos
    Mohammed Hoque, Rana el-Kaliouby, Rosalind Picard

Session 6, 16.00-17.30: Agents in Virtual Worlds and Games

  1. Spontaneous Avatar Behavior for Human Territoriality
    Pedica Claudio, Hannes Vilhjalmsson
  2. Tree Paths: A New Model for Steering Behaviors
    Rafael Rodrigues, Marcelo Paravisi, Alessandro Bicho, Claudio Jung, Leo Pini Magalhaes, Soraia Musse
  3. Getting Acquainted in Second Life - Human Agent Interactions in Virtual Worlds
    Matthias Rehm, Christian Pallay
  4. A virtual tour guide for virtual worlds
    Dusan Jan, Antonio Roque, Anton Leuski, Jacki Morie, David Traum
  5. Design and Evaluation of a Virtual Salesclerk
    Christopher Mumme, Niels Pinkwart
  6. Actors and Characters in Virtual Drama
    Maria Arinbjarnar, Daniel Kudenko


Invited talk 3, 9.00-10.00:
Past and Future Challenges in Creating Emotionally-Engaging Real-Time Digital Actors in Videogames Casey Hudson

Session 7, 10.30-12.00: Tools and Motion Capture

  1. A Realtime Animation Engine for Interactive Embodied Agents
    Alexis Heloir, Michael Kipp
  2. Augmenting Gesture Animation with Motion Capture Data to Provide Full-Body Engagement
    Pengcheng Luo, Michael Kipp, Michael Neff
  3. ION Framework - A Simulation Environment for Worlds with Virtual Agents
    Marco Vala, Guilherme Raimundo, Pedro Sequeira, Pedro Cuba, Rui Prada, Carlos Martinho, Ana Paiva
  4. DTask & LiteBody: Open Source; Standards-based Tools for Building Web-deployed ECAs
    Timothy Bickmore, Daniel Schulman, George Shaw
  5. A Combined Semantic and Motion Capture Database for Real-Time Sign Language Synthesis
    Charly Awad, Nicolas Courty, Kyle Duarte, Thibaut Le Naour, Sylvie Gibet
  6. Mediating Performance Through Virtual Agents
    Gabriella Giannachi, Marco Gillies, Nick Kaye, David Swapp

Session 8 Session 7, 12.00-13.00: Speech and Dialogue

  1. Teaching Computers to Conduct Spoken Interviews: Breaking the Realtime Barrier
    Gudny Jonsdottir, Kristinn Thórisson
  2. Should Agents Speak Like; um; Humans? The Use of Conversational Fillers by Virtual Agents
    Laura Pfeifer, Timothy Bickmore
  3. Turn management or Impression Management?
    Mark Ter Maat, Dirk Heylen
  4. Human-centered Distributed Conversational Modeling: Efficient Modeling of Robust Virtual Human Conversation (video)
    Brent Rossen, Scott Lind, Benjamin Lok

Presenter's Guide

Oral Presentation

Long paper presentations should not exceed 20 minutes, short paper presentations should be limited to 10 minutes. In addition to the presentation times, after each talk there will be time for some questions and discussion.

Poster Presentation

The size of the poster should be A1, material to fix your posters to the boards will be available.

In case you would like to present a small demo of your work on your laptop, please let us know so. In this way we may be able to organize a spot that is more convenient for this.

Finally, and perhaps most importantly, we would like to reward the effort that goes into preparing and presenting a poster by introducing a Best Poster Award. The Best Poster will be selected by public voting.

In case you have any further questions, please don't hesitate to contact me: heylen@ewi.utwente.nl.


The IVA'09 conference is soliciting full papers (12-14 pages), short papers (6-7 pages), and poster papers (1-2 pages). Accepted full length papers will receive a slot for oral presentation in the conference. Accepted short papers will be presented orally or, at the discretion of the conference chairs, presented in the poster session. Poster papers will be presented in the poster session.

All submissions are evaluated in a double-blind peer-review process. The reviewers' main criteria for evaluation are originality, relevance, significance, presentation quality, adequate assessment of the state-of-the-art, and overall quality of the contribution. All accepted papers (full, short, poster) will be published in book form by Springer.

Submission Guidelines

Anonymous submissions should be made electronically in PDF format, and prepared following the formatting instructions of Springer: Information for LNCS Authors (see under Proceedings and Other Multiauthor Volumes). For the sake of anonymity, remove any information from your submission that identifies you or any of the other authors, or any of your institutions or places of work. This year, authors may also submit a companion video (see below).


Submissions will be accepted starting on March 31, 2009.

Papers (all categories) must be submitted by Sunday, April 26, 2009 (11:59pm GMT)

Companion Video

A companion video can underline qualities of your actual system and provide supplementary information. For each paper submission, we accept one companion video (QuickTime, AVI, MPEG or DivX) with a maximum length of 5 minutes and a maximum file size of 100 MB. Please make sure that the video does not reveal the identity of the authors.

Demo Session, Journal Publication, Best Paper Award

All authors of accepted papers will be encouraged to give demo in the demo session (details will follow). Selected papers will be invited for re-submission for a special issue of the Applied AI Journal. As every year, a Best Paper Award will be given by a dedicated jury to the most outstanding conference contribution.


The Gathering of Animated Lifelike Agents (GALA) will take place at IVA-09. For more information about IVA GALA visit please:


The winner of the IVA GALA Awards is Quan Nguyen for his tool Gantool.

The audience prize recieved Kathrin Janowski, Ionut Damian and Dominik Sollfrank for their tennis spectators.

All the videos can be found here.




Early registration can be done until June 30!

For registration, including rates, please go to:

IVA registration website

Full Registration includes a conference bag with proceedings and maps and info about Amsterdam, two lunches and a reception at NEMO as well as drinks during coffee breaks. Furthermore registration entitles participants to reserve a reduced-fee ticket for the social dinner at the Amsterdam Zoo, with a surprise entertainment program. Reservation for the dinner can be made on the registration form.


For further information send your inquiries to iva09@dfki.de.

Important Dates

Papers and Posters


Conference Chairs

Best Paper Chair

Submissions Chair

Poster and Demo Chair

GALA Chair

Local Organization Chair

Senior Program Committee

Program Committee