arXiv (Cornell University) · 2017 · 31 citations · 0 references
ComputingComputer ScienceArtificial IntelligenceMachine LearningMotor ControlDeep LearningGame PlayingMulti-agent SystemsPolicy LearningReinforcement LearningArtificial Intelligence In GamesDeep Reinforcement LearningGame-playing AgentsGeneric AlgorithmsLearning From PlayMulti-agent LearningPartial ObservabilityPartial Observability Models
Game‑playing AI has recently advanced, with deep‑learning RL algorithms solving diverse tasks from Atari to board games, yet the complex, partially observable, multiplayer dynamics of Super Smash Bros. Melee remain a challenging domain for both humans and machines. The study investigates whether deep‑learning RL methods can master Super Smash Bros. Melee and demonstrates that agents can compete with and surpass professional human players.
There has been a recent explosion in the capabilities of game-playing artificial intelligence. Many classes of RL tasks, from Atari games to motor control to board games, are now solvable by fairly generic algorithms, based on deep learning, that learn to play from experience with minimal knowledge of the specific domain of interest. In this work, we will investigate the performance of these methods on Super Smash Bros. Melee (SSBM), a popular console fighting game. The SSBM environment has complex dynamics and partial observability, making it challenging for human and machine alike. The multi-player aspect poses an additional challenge, as the vast majority of recent advances in RL have focused on single-agent environments. Nonetheless, we will show that it is possible to train agents that are competitive against and even surpass human professionals, a new result for the multi-player video game setting.