Introduction

On Monday November 13th, 2017, from 9AM to 6PM, a satellite workshop "Investigating Social Interactions with Artificial Agents" is organized in parallel to the 19th ACM International Conference on Multimodal Interactions in Glasgow by the Institute for Langage, Communication and the Brain. In addition to four series of talks (artificial agents, linguistique, social signal processing and neuroscience), attendees are invited to present posters or demos of their work before July 20th, 2017. A two-page presentation following ACM templates (link to ACM templates) by following the link "Submission" on the left of the page. Attendance is free, but registration (see link "Register" on the left) mandatory for organisational purposes. 

Objectives

There is a growing need for understanding the changes caused by social interactions in realistic settings, from pairs to large groups and crowds at the behavioural, neural and physiological levels. Based on this, this workshop seeks to review how new technologies for interacting artificial agents can be and are being used to increase our knowledge deeper social interaction involving, for instance, emotion, and factors of embodiment and cognition. Realistic interaction contexts require new experimental tools and paradigms, combining social sciences and humanities with neuroscience, engineering and computing. This is necessary to ground future social media with the aim of improving social engagement, mutual understanding, collaboration, cognitive resonance, with outputs in entertainment and wellbeing, among others.

The aim of the workshop is to bring together an interdisciplinary group of experts in three complementary fields. From humanities, the core of research in linguistics and social sciences has been to investigate behavioural and cognitive underpinnings of natural social interactions through speech. In neuroscience, investigating natural social interactions with modern neuroimaging techniques is the new frontier that must be reached to truly understand the neural specificities of normal as well as pathological human behaviours. Computer science has made tremendous progresses in the extraction of social signal from multimodal data, including audio and video recordings of natural interactions. In addition, computer science allow us to close the loop, by providing controllable artificial agents – such as computer animated agents or humanoid robots – that can be used as benchmarks to test hypotheses about behaviours influencing the social competence of an agent.

Altogether, efforts must be – and are being – made to boost this tryadic interaction between computer science, humanities and neurosciences. The aim of this workshop will be to illustrate with past and currently ongoing projects, the fruits that such endeavours can bear. In addition to boosting existing interactions, the objective of the workshop will be promote a larger community of researchers endorsing this approach. 

Program (October 25th, 2017)

*****************************************************************
Program of ICMI'17 satellite workshop  "Investigating Social Interactions with Artificial Agents," organized by T. Chaminade, M. Ochs, N. Nguyen and F. Lefèvre.

 9.00  General introduction (organizers)


o Neuroscience of social cognition using artificial agents (chair T. Chaminade).
 9.15   Thierry Chaminade, INT CNRS - Aix-Marseille Univ, FR - How do artificial agents think?
 9.30   Marc Cavazza, University of Kent, Canterbury, UK - Brain-Computer Interfacing to Social Agents.
10.10   Agnieska Wykowska, Istituto Italiano di Tecnologia, Genova, IT - Using embodied artificial agents to understand human social cognition.

o 10.50 Coffee Break

o Automatic analysis of social signals (chair M. Ochs).
11.00   Alessandro Vinciarelli, Glasgow Univ., UK - Body Language without a Body: Nonverbal Communication in Technology Mediated Settings.
11.40   Louis-Philippe Morency, School of Computer Science, Carnegie Mellon University, USA - Multimodal Machine Learning: Modeling Human Communication Dynamics.

o 12.20 Lunch

o Endowing artificial agents with spoken language skills (chair N. Nguyen).
 1.30   Philippe Blache, LPL, Aix-en-Provence, FR - Dialogue Management in Task-Oriented Dialogue Systems.
 2.10   Matthew Aylett, School of Informatics, The University of Edinburgh, UK - Bot or Not? Exploring the perception of acted, modified and synthetic speech.

o 2.50  Poster session (and Coffee Break)

o Control of social artificial agents (chair F. Lefèvre).
 4.00   Catherine Pelachaud, ISIS, FR - Greta: A Conversing Socio-emotional Agent.
 4.40   Louisa Pragst, Ulm Univ., GE - Challenges for Adaptive Dialogue Management in the KRISTINA Project.
 5.20   Fabrice Lefèvre, LIA, Univ. Avignon, FR - En Route to a Better Integration and Evaluation of Social Capacities in Vocal Artificial Agents.

o 5.50 General Discussion
*****************************************************************

Organizers

Institute for Langage, Communication and the Brain, ilcb.fr

The ILCB constitutes a major initiative in Cognitive Science. It aims to explore the neural basis of human language and communication by studying the components of L&C and proposing unifying models and architectures. The originality and ambition of the ILCB lies in the integration of a wide range of disciplines, methodologies and experimental techniques that all contribute to the description of human language processing and communication (as paradigmatic examples of high-level cognitive functions) in laboratory and naturalistic settings.

Thierry Chaminade, Institut de Neurosciences de la Timone (INT), CNRS – Aix-Marseille Université, thierry.chaminade@univ-amu.fr

Thierry Chaminade is a French tenure CNRS researcher in the field of neuroscience. His scholarly interest is the biological bases of the uniqueness of the human mind. In practice, different measures of human physiology, from biochemistry to neuroimaging, are combined with approaches from humanities, from anthropology to psychology and psychiatry, in order to develop an integrative account of human cognition. His research investigates the response to the emergence of artificial agents, computer-animated avatars and humanoid robots, to assess how humans adapt their social skills in the absence of pre-existing concepts about the nature and behaviour of these agents. He is a member of the Institute for Language, Communication and the Brain at Aix-Marseille.

Noël Ngyuen, Laboratoire Parole & Langage, CNRS – Aix-Marseille Université, noel.nguyen-trong@univ-amu.fr

Noël Nguyen is Professor (Classe exceptionnelle) in Speech and Language Sciences at Aix-Marseille University, Head of the Laboratoire Parole et Langage (CNRS & Aix-Marseille University), and Vice-Dean for Research, Faculty of Arts, Letters, Languages, and Human Sciences, at Aix-Marseille University. He is also a former member of the French National Committee for Scientific Research. Noël Nguyen's research focuses on the production, perception and comprehension of spoken language. He was or is the coordinator of several projects conducted under the auspices of the French Ministry of Research, the French National Center for Scientific Research, the French National Agency for Research, the Brain and Language Research Institute at Aix-Marseille. He was the scientist-in-charge for the LPL within the Sound-to-Sense Research and Training Marie Curie European Network. He currently coordinates the COBRA (Conversational Brains) international research network. He is a member of the Institute for Language, Communication and the Brain at Aix-Marseille.

Magalie Ochs, Laboratoire Sciences de l’Information et des Systèmes (LSIS), CNRS – Aix-Marseille Université, magalie.ochs@lsis.org

Magalie Ochs is Associate Professor in Computer Science at Aix-Marseille University in the LSIS laboratory. She has explored different computational methods and models to endow virtual characters with socio-emotional capabilities (reasoning and expression). A recent interest aims at exploiting corpus-based machine learning methods to construct stochastic and graphical model to compute in real-time virtual agent’s social signals. She is a member of the Institute for Language, Communication and the Brain at Aix-Marseille.

Fabrice Lefevre, Laboratoire d’Informatique d’Avignon, Université d’Avignon et des Pays de Vaucluse, fabrice.lefevre@univ-avignon.fr

Fabrice Lefèvre received a degree in electrical engineering from ENSEA-Cergy and the Ph.D. degree in computer science from the University Paris VI, Paris, France, in 2000. He was appointed an Assistant Professor position at the University of Orsay, Paris XI, in 2001 where he worked in the Spoken Language Processing Group at LIMSI-CNRS. He joined the University of Avignon in 2005, and he is now Professor and head of the Vocal Interactions Group at LIA since 2010. His primary research activities include automatic speech recognition, speech understanding, stochastic machine translation and spoken dialog systems. He is a member of the Institute for Language, Communication and the Brain at Aix-Marseille.

Online user: 1