UE937 - COGSCI 314 - Robotic modelling approaches to Cognitive Sciences

Type d'UE
Méthodologie
Disciplines
Psychologie et sciences cognitives
Page web
https://cogmaster.ens.psl.eu/en/program/m2-program-13572 
Langues
anglais
Mots-clés
Modélisation Sciences cognitives
Aires culturelles
-

The purpose of this course is to present recent work in cognitive robotics and how it can contribute to fundamental questions in cognitive sciences. Historically tightly related to AI, cognitive robotics puts a strong emphasis on how high-level reasoning and cognition are anchored on low-level sensorimotor processes and on embodiment (the role of the body in cognition). Experimenting with robots also forces one to make hypotheses about how different cognitive functions interact (perception, decision, action, control motor, exploration / curiosity, learning, social interaction) and should be integrated within a cognitive architecture in order to make things work in the real world. The course will present advances in robot learning (incl. deep learning), Bayesian approaches, evolutionary robotics, navigation, human-robot interaction, optimal control, and developmental learning. Links with computational neuroscience & developmental psychology will be shown.

- Introduction: Towards robotics models of cognitive architectures (Mehdi Khamassi / Raja Chatila)

- Active perception and its links with Kevin O’Regan’s sensorimotor theory (Bruno Gas)

- Discrete and deep reinforcement learning (Olivier Sigaud)

- Model-based/model-free neurorobotics reinforcement learning models (Mehdi Khamassi)

- Deep learning (Charles Ollion)

- Developmental robotics and artificial curiosity (Pierre-Yves Oudeyer)

- Spatial cognition and navigation (Angelo Arleo)

- Coordination of navigation strategies and reinforcement learning (Benoît Girard)

- Bayesian approaches to robotics (Pierre Bessière)

- Social signals, imitation and emotion for interactions (Mohamed Chetouani / Philippe Gaussier)

- Motor control (Emmanuel Guigon)

- Evolutionary robotics (Nicolas Bredèche / Stéphane Doncieux)

- Language-based human-robot interaction (Peter Dominey)

Practical labs:

Discrete model-based/model-free reinforcement learning (Nicolas Perrrin)

Multi-strategy navigation in simulation (Benoît Girard)

  • Sciences cognitives – M2/S3
    Suivi et validation – semestriel hebdomadaire = 6 ECTS
    MCC – exposé oral, autre (TP)
  • Mehdi Khamassi [référent·e]   chargé de recherche, CNRS /
Contacts additionnels
cogmaster@psl.eu
Informations pratiques

The complete syllabus of the course is available on the Cogmaster's website. For any information, please contact the secretariat of the Cogmaster.

Registration procedure (external students) : https://cogmaster.ens.psl.eu/en/students/external-students-13501

Direction de travaux des étudiants
-
Réception des candidats
-
Pré-requis

Good knowledge in maths & programming.

  • Autre lieu Paris
    45 rue des Saints-Pères 75006 Paris (salle à préciser)
    1er semestre / hebdomadaire, vendredi 09:00-12:00
    du 25 septembre 2020 au 22 janvier 2021


Intervenant·e·s


  • Mehdi Khamassi [référent·e]   chargé de recherche, CNRS /

Planning


  • Autre lieu Paris
    45 rue des Saints-Pères 75006 Paris (salle à préciser)
    1er semestre / hebdomadaire, vendredi 09:00-12:00
    du 25 septembre 2020 au 22 janvier 2021


Description


Type d'UE
Méthodologie
Disciplines
Psychologie et sciences cognitives
Page web
https://cogmaster.ens.psl.eu/en/program/m2-program-13572 
Langues
anglais
Mots-clés
Modélisation Sciences cognitives
Aires culturelles
-

The purpose of this course is to present recent work in cognitive robotics and how it can contribute to fundamental questions in cognitive sciences. Historically tightly related to AI, cognitive robotics puts a strong emphasis on how high-level reasoning and cognition are anchored on low-level sensorimotor processes and on embodiment (the role of the body in cognition). Experimenting with robots also forces one to make hypotheses about how different cognitive functions interact (perception, decision, action, control motor, exploration / curiosity, learning, social interaction) and should be integrated within a cognitive architecture in order to make things work in the real world. The course will present advances in robot learning (incl. deep learning), Bayesian approaches, evolutionary robotics, navigation, human-robot interaction, optimal control, and developmental learning. Links with computational neuroscience & developmental psychology will be shown.

- Introduction: Towards robotics models of cognitive architectures (Mehdi Khamassi / Raja Chatila)

- Active perception and its links with Kevin O’Regan’s sensorimotor theory (Bruno Gas)

- Discrete and deep reinforcement learning (Olivier Sigaud)

- Model-based/model-free neurorobotics reinforcement learning models (Mehdi Khamassi)

- Deep learning (Charles Ollion)

- Developmental robotics and artificial curiosity (Pierre-Yves Oudeyer)

- Spatial cognition and navigation (Angelo Arleo)

- Coordination of navigation strategies and reinforcement learning (Benoît Girard)

- Bayesian approaches to robotics (Pierre Bessière)

- Social signals, imitation and emotion for interactions (Mohamed Chetouani / Philippe Gaussier)

- Motor control (Emmanuel Guigon)

- Evolutionary robotics (Nicolas Bredèche / Stéphane Doncieux)

- Language-based human-robot interaction (Peter Dominey)

Practical labs:

Discrete model-based/model-free reinforcement learning (Nicolas Perrrin)

Multi-strategy navigation in simulation (Benoît Girard)


Master


  • Sciences cognitives – M2/S3
    Suivi et validation – semestriel hebdomadaire = 6 ECTS
    MCC – exposé oral, autre (TP)

Renseignements


Contacts additionnels
cogmaster@psl.eu
Informations pratiques

The complete syllabus of the course is available on the Cogmaster's website. For any information, please contact the secretariat of the Cogmaster.

Registration procedure (external students) : https://cogmaster.ens.psl.eu/en/students/external-students-13501

Direction de travaux des étudiants
-
Réception des candidats
-
Pré-requis

Good knowledge in maths & programming.