Fighting Game

Fighting games, encompassing both human-versus-human and human-versus-AI gameplay, are a rich domain for studying artificial intelligence and human-computer interaction. Current research focuses on developing robust and adaptable AI agents using deep reinforcement learning (DRL), often employing transformer networks or proximal policy optimization (PPO), to improve agent performance, generalizability across characters, and alignment with human player expectations. These advancements are not only improving the AI opponents in commercial games but also providing valuable insights into efficient data management, adaptive game design (such as dynamic music), and the development of more human-like AI behaviors. The resulting improvements in AI agents and game design contribute to a deeper understanding of complex decision-making processes and enhance the overall gaming experience.