Google Research Football: A Novel Reinforcement Learning Environment
This provides a new, open-source environment for reinforcement learning researchers to test algorithms in a safe and reproducible way, though it is incremental as it builds on existing virtual environment concepts.
The authors tackled the need for a challenging and customizable reinforcement learning environment by introducing the Google Research Football Environment, a physics-based 3D simulator for training agents to play football, and reported baseline results for three algorithms (IMPALA, PPO, and Ape-X DQN) on proposed benchmarks.
Recent progress in the field of reinforcement learning has been accelerated by virtual learning environments such as video games, where novel algorithms and ideas can be quickly tested in a safe and reproducible manner. We introduce the Google Research Football Environment, a new reinforcement learning environment where agents are trained to play football in an advanced, physics-based 3D simulator. The resulting environment is challenging, easy to use and customize, and it is available under a permissive open-source license. In addition, it provides support for multiplayer and multi-agent experiments. We propose three full-game scenarios of varying difficulty with the Football Benchmarks and report baseline results for three commonly used reinforcement algorithms (IMPALA, PPO, and Ape-X DQN). We also provide a diverse set of simpler scenarios with the Football Academy and showcase several promising research directions.