I am research director (DR1) at Inria, heading the Flowers team at Inria Bordeaux Sud-Ouest (see PhD students). I was previously a permanent researcher in Sony Computer Science Laboratory for 8 years (1999-2007).
I have been studying lifelong autonomous learning, and the self-organization of behavioural, cognitive and cultural structures, at the frontiers of artificial intelligence, machine learning, cognitive sciences and educational technologies.
In particular, I study mechanisms of exploration in autonomous learning, with a special focus on mechanisms enabling agents to set their own goals, and how this can self-organize curriculum learning. This includes mechanisms of intrinsically motivated learning (also called curiosity-driven active learning), autonomous unsupervised exploration, imitation and social learning, multimodal statistical inference, embodiment and maturation and self-organization.
This developmental machine learning work was recently integrated with deep reinforcement learning techniques, where unsupervised deep learning techniques are used to learn spaces in which to self-generate goals, and where curiosity-driven exploration is used to discover independently controllable features and solve efficiently sparse reward problems in Deep RL.
I consider cognitive development as a complex dynamical system which needs to be understood through systemic thinking, leveraging tools and concepts from computational sciences (artificial intelligence, machine learning and robotics), neuroscience and psychology. I consider algorithmic models as powerful scientific languages to express theories of cognitive development in the living.
I am also working on applications of this research in three fields: adaptive human-computer interfaces, educational technologies and open-source robotics for art and education.
In practice, this has led me to follow two mutually reinforcing research activities (yet with two different epistemological perspectives):
Link to information about the Poppy humanoid robot : Poppy Project web site. Poppy is an open-source 3D printed robot for science, education and art designed by the Flowers team. It was built to study the impact of the body on sensorimotor development and cognition: it makes it possible to really consider the body as an experimental variable. See article at Humanoids 2013 conference. Poppy Overview from Poppy Project on Vimeo. I collaborate regularly with artist within project that explore the frontiers between art and science. This has been the opportunity to create original connections between the general public and our scientific projects, in particular by bringing people to ask themselves and to ourselves stimulating questions about the position of such scientific projects within society at large. Examples of such projects include:
I will receive in october 2018 the Inria Prize of the National Science Academy (computer science researcher under 40).
Laversanne-Finot, A., Péré, A., Oudeyer, P-Y. (2018) Curiosity Driven Exploration of Learned Disentangled Goal Spaces, In Proceedings of Conference on Robot Learning (CoRL 2018). Oudeyer, P-Y. (2018) Computational Theories of Curiosity-driven Learning, in “The New Science of Curiosity”, ed. G. Gordon, NOVA. Colas, C., Sigaud, O., Oudeyer, P. Y. (2018). GEP-PG: Decoupling Exploration and Exploitation in Deep Reinforcement Learning Algorithms. In Proceedings of International Conference on Machine Learning (ICML 2018), arXiv preprint arXiv:1802.05054. Other links: code and blog post. Péré, A., Forestier, S., Sigaud, O., & Oudeyer, P. Y. (2018). Unsupervised Learning of Goal Spaces for Intrinsically Motivated Goal Exploration. In Proceedings of International Conference on Learning Representations (ICLR 2018), arXiv preprint arXiv:1803.00781. Other links: code. I gave the 29th Gibson Lecture at University of Cornell, Ithaca. The talk was about “Active exploration mechanisms in autonomous learning and development” (slides), and in particular I presented several robotic and machine learning models of automatic intrinsically motivated curriculum learning. The demonstration “Intrinsically motivated multitask reinforcement learning” built by the Flowers team was runner up of the NIPS 16 demo competition. This demonstration shows real time online acquisition of repertoires of high-dimensional nested skills (including tool use) with curiosity-drive goal exploration processes. Together with J. Gottlieb, M. Lopes, and C. Kidd, we have been awarded a 3 year grant by the HFSP program for an interdisciplinary research program to study active exploration and learning, and their computational modeling, in infants and monkeys. We are looking for motivated PhD and postdoc candidates to work on this project. Together with J. Gottlieb and T. Gliga, we have co-organized the Second Interdisciplinary Symposium on Information Seeking, Curiosity and Attention in London, in september 2016. This great event has gathered researchers from neuroscience, psychology and computer science/machine learning: J. Nelson, D. Markant, R. Ligneul, S. Kouider, M. Gruber, K. Murayama, J. O’Reilly, G. Baldassarre, P. Dayan, P-Y. Oudeyer, K. Doya, W. Shultz, A. Bell, L. Hunt, J. Gottlieb, D. Bell, K. Begus, L. Goupil, L. Feigenson, D. Bavelier, T. Gliga. Most videos and slides of the symposium are available on the Neurocuriosity 2016 symposium web site. Poppy project is an open-source platform for the creation, use and sharing of interactive 3D printed robots. It was initially developed by the Flowers team at Inria, and is now supported by a large community of contributors and users. The new website of Poppy Project is out, and provides all links to hardware, software and documentation resources to build and/or obtain an open-source 3D printed Poppy robot, under the Creative Commons licence (CC-BY-SA). Poppy Humanoid was the first worldwide open-source 3D printed full body humanoid robot. Poppy Torso and Poppy Ergo are other robots that can be built and modified with the Poppy platform. The Poppy platform is used widely for projects in open science, open education from primary schools to university, and art. Join the interdisciplinary community on the forum, and explore the GitHub repositories. See the Jobs page on the Flowers website.Cognitive Science: Models of Human Development and Language Evolution
ICDL-Epirob – International Conference on Development and Learning, Epirob, Oct 2014, Genoa, Italy.
Machine Learning, Artificial Intelligence, Deep RL and Developmental Robotics,
. (2017)
arXiv:1708.02190
ICDL-Epirob – International Conference on Development and Learning, Epirob, Oct 2014, Genoa, Italy.
Adaptive Human-Machine Interfaces, BCI, Emotional Speech Processing
; Fabien Danieau; David Filliat (2013) The Impact of Human-Robot Interfaces on the Learning of Visual Objects ,
IEEE Transactions on Robotics, IEEE Robotics and Automation Society, 2013, 29 (2), pp. 525-541 Bibtex
Educational Technologies
Outreach Projects
Poppy: an Open-Source 3D Printed Robotic Platform
IniRobot: un kit pédagogique pour l'initiation à la robotique à l'école primaire
IniRobot: IniRobot est une série d’activités pédagogiques “clés en main” destinée à la découverte de la robotique et de la programmation à l’école primaire, en particulier lors des activités périscolaires. Ce kit est libre d’utilisation (Creative Commons CC-BY-SA) et utilise le robot Thymio développé à l’EPFL. Il est déployé en France dans les activités périscolaires des écoles de plusieurs villes, dont Lille, Talence et Lormont. Il a été développé par Didier Roy, Thomas Guitard et Pierre-Yves Oudeyer dans l’équipe Flowers, et est partagé sur le site participatif Dessine-moi un robot.
Art and Science
News
Inria/National Academy of Science Prize 2018
New papers on curiosity, deep RL and lifelong learning
Gibson Lecture at Cornell University (04/17)
NIPS 2016 Demo awards: Flowers team demo was runner up
Human Frontier Science Program grant Curiosity accepted (March 2016)
Second Interdisciplinary Symposium on Information Seeking, Curiosity and Attention, Sept. 2016
Poppy Project: Open-Source 3D printed Robotics Creation platform for Science, Education and Art
Jobs
Mondes Mosaiques: Astres, villes, vivant et robots (CNRS Editions, 2015)
Aux sources de la parole (Odile Jacob, 2013)
Self-organization in the evolution of speech (Oxford University Press, 2006)
The Economist, sept. 2018: A sense of curiosity is helpful for artificial intelligence; Scientific American, 2018: Intelligent Machines that Learn Like Children; Pour la Science, 2015: L’éveil des bébés robots La recherche, 2015 Le roboticien des sciences humaines Le Monde, nov. 2014 (portrait): Pierre-Yves Oudeyer, aussi curieux que ses robots Socialter, sept. 2014: Les robots seront-ils aussi “bêtes” que nous? Des machines et des hommes Les Echos, mars 2014 : Les robots auront un impact sur la société
Vidéo Lift: “Quand les robots nous aident à comprendre l’homme” , avec présentation de Poppy, robot humanoide open-source et imprimé en 3D.
France Culture (oct. 2013) La parole et l’ordinateur, Interview avec Stéphane Delogeorges, émission Continent Sciences
France Inter (oct. 2013) Le langage: une auto-organisation ? Interview avec Stéphane Paoli, émission 3D, le journal
RFI (Sept. 2013) Comment s’invente le langage ? Emission “Autour de la question” de Caroline Lachowsky.
L’expérience Ergo-Robots, expostion “Mathématiques un Dépaysement Soudain”, Fondation Cartier, Paris