Etienne de Sevin's Research Page

Octavia Project (National: FSN) 2013 - 2016 (Link)

OCTAVIA (Outil de Création et Test d’Ambiance de Vie Interactive et Autonome) is a research project funded by PIA/FSN (Projet Investissement d’avenir/Fonds nationale pour la Société Numérique). This project is labeled by the French “Cap Digital” business cluster for digital content and services.

The main goal of the project is to propose an innovative tool to design and test scenes of ambient life involving interactive and autonomous non-player character in 3D environment. The project proposes other outcomes : a new idea of 3D virtual cities proposing an environment populated with small groups of NPCs that play interactive scenarios to increase the presence of the user in the environment ; a videogame that use interactive and autonomous groups of NPCs to propose an innovative gameplay experience ; a sound engine coupled with 3D animation and environment to improve immersion.

Industrial and research partners: MASA Group (Coordinator), BSE Group, Black Sheep Studio, CNAM (ENJMIN)

Terra Dynamica Project (National: FUI) 2010 - 2013 (Link) (Video)

One of the main motivations coming from applications for working on artificial general intelligence is the design of agent architectures which are sufficiently generic to apply to a wide range of application domains. We design a generic architecture for artificial intelligence for agents in urban simulations with various forms of intelligence.

The originality of our approach is that it deals, at a high level,with affects, logical reasoning and coordination, and arbitrates between the candidate behaviors originating from all these modules.It is currently being developed in our laboratory within the context of Terra Dynamica, a large collaborative project aiming for an AI framework for rich multi-agent urban simulations. Each high-level module has its own coherence and will contribute to a versatile simulation framework. In parallel, we have addressed the issue of scalability so that solutions outlined here can handle several thousands of agents.

Industrial and research partners: Thales (Coordinator), Lip6 - Université Pierre et Marie Curie, IFSTTAR, CITU-Paragraphe -Université Paris 8, Laboratoire Cedric - CNAM, Star-Apic, BeTomorrow, Kylotonn, Davi

SEMAINE Project (European) 2008 - 2011 (Link) (Video)

The Semaine project is an EU-FP7 1st call STREP project and aims to build a SAL, a Sensitive Artificial Listener, a multimodal dialogue system which can:

interact with humans with a virtual character
sustain an interaction with a user for some time
react appropriately to the user's non-verbal behaviour

The aim of the SEMAINE project is to build a Sensitive Artificial Listener : a multimodal dialogue system with the social interaction skills needed for a sustained conversation with a human user. The system will emphasise communication skills, i.e. non-verbal, social and emotional perception, interaction and behaviour capabilities. The Sensitive Artificial Listener paradigm involves only very limited verbal capabilities, but has been shown to be suited for prolonged human-machine interaction. In this paradigm, we will build a real-time, robust interactive system perceiving a human user's facial expression, gaze, and voice, and engaging with the user through an Embodied Conversational Agent's body, face and voice. The agent will exhibit audiovisual listener feedback in real time while the user is speaking, and will take the user's feedback into account while the agent is speaking. The agent will pursue different dialogue strategies depending on the user's state; it will learn to interpret the user's non-verbal behaviour and adapt its own behaviour accordingly.

In the end, this SAL-system will be released to a large extent as an open source research tool to the community.

Industrial and research partners: DFKI Speech Group - German Research Center on Artificial Intelligence - Germany, i·BUG group - Imperial College London - UK, Great Team - Telecom ParisTech - France, Perception Action Communication - Queen's University - Belfast - UK, Machine Intelligence and Signal Processing (MISP) Group - Munich's University of Technology -Germany, The Human Media Interaction University of Twente - Holland

HUMAINE Project (European) 2004 - 2008 (Link)

Emotion-oriented computing is a broad research area involving many disciplines. The EU-funded network of excellence HUMAINE (Human-Machine Interaction Network on Emotion) is currently making a co-ordinated effort to come to a shared understanding of the issues involved, and to propose exemplary research methods in the various areas.

HUMAINE aims to lay the foundations for European development of systems that can register, model and/or influence human emotional and emotion-related states and processes - 'emotion-oriented systems'. Such systems may be central to future interfaces, but their conceptual underpinnings are not sufficiently advanced to be sure of their real potential or the best way to develop them. One of the reasons is that relevant knowledge is dispersed across many disciplines. HUMAINE brings together leading experts from the key disciplines in a programme designed to achieve intellectual integration. It identifies six thematic areas that cut across traditional groupings and offer a framework for an appropriate division of labour - theory of emotion; signal/sign interfaces; the structure of emotionally coloured interactions; emotion in cognition and action; emotion in communication and persuasion; and usability of emotion-oriented systems. Teams linked to each area will run a workshop in it and carry out joint research to define an exemplar embodying guiding principles for future work in their area. Cutting across these are plenary sessions where teams from all areas report; activities to create necessary infrastructure (databases recognising cultural and gender diversity, an ethical framework, an electronic portal); and output to the wider community in the form of a handbook and recommendations of good practice (as precursors to formal standards).

Industrial and research partners: 33 partners from 14 countries participate in the network.