Personnal Page

Seurin Mathieu

Phd Student in Deep Reinforcement Learning at Sequel (CRIStAL/Inria Lab in Lille - France)

Hi, I'm Mathieu SEURIN

I'm P.h.D Student working on Deep Reinforcement Learning. I did a Master's Degree in Machine Learning, Big Data and Knowledge Representation (DAC) at the french Université Pierre et Marie Curie.

Currently under the supervision of Pietquin Olivier (Google Brain) and Preux Phillipe (team SequeL at CRIStAL/INRIA), i'll be focusing on Deep Reinforcement Learning applied to dialogue systems during the thesis.

I'm interested in Reinforcement Learning, Deep Learning, Multi-Modal Learning, Multi-goal/task/continual learning, dialog system, Robotics and to another extent, Music.

Publications (Google scholar)

  • IJCAI'21 : Don't Do What Doesn't Matter : Instrinsic Motivation with Action Usefulness (M. Seurin, F. Strub, P. Preux, O. Pietquin) https://arxiv.org/abs/2105.09992

  • Interspeech'20 : A Machine of Few Words - Interactive Speaker Recognition with Reinforcement Learning (M. Seurin, F. Strub, P. Preux, O. Pietquin) https://arxiv.org/abs/2008.03127

  • IJCNN'20 : "I'm sorry Dave, i'm afraid i can't do that" Deep Q-Learning from forbidden action (M. Seurin, P. Preux, O. Pietquin) https://arxiv.org/pdf/1910.02078.pdf

  • APDRL'20 : HIGhER : Improving instruction following with Hindsight Generation for Experience Replay (G. Cideron*, M. Seurin*, F. Strub, O. Pietquin) https://arxiv.org/abs/1910.09451

  • IJCNN'19 : Unsupervised state representation learning with robotic priors: a robustness benchmark (T. Lesort*, M. Seurin*, X. Li, N. Díaz Rodríguez, D. Filliat) http://arxiv.org/abs/1709.05185

  • ECCV'18 : Visual Reasoning with a Multi-hop FiLM Generator (F. Strub, M. Seurin, E. Perez, H. De Vries, J. Mary, P. Preux, A. Courville, O. Pietquin) https://arxiv.org/abs/1808.04446

* = Equal Contribution.

Event/Organization

[2019] Co-Organizer of Reinforcement Learning Summer School (RLSS) with other SequeL member https://rlss.inria.fr/

[2018] Co-Organizer of the European Workshop on Reinforcement Learning (EWRL) with Matteo Pirotta, Ronan Fruit and Florian Strub https://ewrl.wordpress.com/ewrl14-2018/

Talk

Template can be found here : https://github.com/eleurent/inria-beamer

[10/2019] Sequel Seminar : "I'm Sorry Dave, I'm afraid I can't do that ... I didn't understand" (slides available)

Contact :

firstname.lastname@inria.fr

Github,

Courses :

I also give courses to the Lille3's student in Machine Learning, Python, Data-Processing and Database Management