Jump to Navigation

Robots Exploring with Curiosity

    Hello !

    I'm a third year PhD student at INRIA Bordeaux under the supervision of Pierre-Yves Oudeyer.

    Contact me at sebastien.forestier@inria.fr



My topics of interests include developmental psychology and developmental robotics:  we try to create curious robots that learn like children, both to inform, make hypothesis and predictions on children learning mechanisms, and to come up with robots better able to learn in complex child-like environments. In particular, we are interested in the development of tool-use and vocal skills in children and we study the development of those skills in robots with curiosity-driven learning mechanisms and social feedback.

On this website you will find videos and publications about those projects.


Open Source Software:

I'm one of the maintainers of the Explauto Python library allowing to study, model and simulate curiosity-driven learning and exploration in virtual and robotic agents. Take a look !

We also share on Github all the Python code of the experiments we described in papers, e.g. in the repositories NIPS2016, CogSci2016, ICDL2016, and IROS2016, Intrinsically Motivated Goal Exploration Processes with Automatic Curriculum Learning.



You can also find a tutorial on the exploration algorithms we developed with comparisons of several concepts (Motor Babbling, Goal Babbling, Model Babbling). This Jupyter notebook can either be read-only, or tested interactively (so that you can study new parameters) if you download the Github repository and run the notebook with Jupyter.


Recent activities:

Main menu 2

Page | by Dr. Radut