责任导师 : Amal EL FALLAH SEGHROUCHNI

助理责任导师 : 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.

2021-2023 刊物