LE PELLETIER DE WOILLEMONT Pierre
Supervision : Amal EL FALLAH SEGHROUCHNI
Co-supervision : Vincent CORRUBLE
Simulation Of Diverse Human Play-Styles In Video Games: A Reinforcement Learning Approach
The increasing complexity of gameplay mechanisms in modern video games is leading to the emergence of a wider range of ways to play games. The variety of possible play-styles needs to be anticipated by designers, through automated tests. In this dissertation, we explore the application of reinforcement learning techniques to create human-like agents that can simulate diverse play-styles for use in automated game testing. We investigate how RL algorithms can be used to generate agents that mimic the behavior of human players and how these agents can be used to evaluate the game. Our objective is to bridge the gap between the needs and constraints of the video game industry and the capabilities and requirements of the existing machine learning algorithms. We developed three approaches, relying on goal-conditioning, imitation learning and adversarial learning. These methods may be used at different stages of the video game production for various purposes.
Defence : 09/11/2023
Jury members :
Stuart RUSSELL, Professeur des universités, University of California [Rapporteur]
Georgios YANNAKAKIS, Professeure des universités, Università ta' Malta [Rapporteur]
Geber RAMALHO, Professeure des universités, UFPE
Olivier SIGAUD, Professeur des universités, Sorbonne Université
Amal EL FALLAH SEGHROUCHNI, Professeure des universités, Sorbonne Université
Vincent CORRUBLE, Maître de conférences, Sorbonne Université
Rémi LABORY, Ubisoft Paris
2021-2023 Publications
-
2023
- P. Le Pelletier de Woillemont : “Apprentissage par renforcement appliqué à la simulation de styles de jeu humains variés dans les jeux vidéos”, thesis, phd defence 09/11/2023, supervision El fallah seghrouchni, Amal, co-supervision : Vincent, CORRUBLE (2023)
-
2022
- P. Le Pelletier de Woillemont, R. Labory, V. Corruble : “Automated Play-Testing Through RL Based Human-Like Play-Styles Generation”, AIIDE'22: Proceedings of the Eighteenth AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment n°18, AIIDE'22, Ponoma, CA, United States, pp. 146-154, (AAAI Press), (ISBN: 978-1-57735-877-0) (2022)
-
2021
- P. Le Pelletier de Woillemont, R. Labory, V. Corruble : “Configurable Agent With Reward As Input: A Play-Style Continuum Generation”, 2021 IEEE Conference on Games (CoG), Copenhagen (Virtual), Denmark, pp. 1-8, (IEEE), (ISBN: 978-1-6654-3886-5) (2021)