Marathon Environments: Multi-Agent Continuous Control Benchmarks in a Modern Video Game Engine
This provides game developers and researchers with accessible tools for multi-agent continuous control research in a modern game engine, though it is incremental as it adapts existing methods to a new platform.
The authors tackled the lack of open-source continuous control benchmarks in commercial game engines by creating Marathon Environments, a suite implemented in Unity that successfully reproduces advanced locomotion tasks like walking, running, and backflipping from motion capture data.
Recent advances in deep reinforcement learning in the paradigm of locomotion using continuous control have raised the interest of game makers for the potential of digital actors using active ragdoll. Currently, the available options to develop these ideas are either researchers' limited codebase or proprietary closed systems. We present Marathon Environments, a suite of open source, continuous control benchmarks implemented on the Unity game engine, using the Unity ML- Agents Toolkit. We demonstrate through these benchmarks that continuous control research is transferable to a commercial game engine. Furthermore, we exhibit the robustness of these environments by reproducing advanced continuous control research, such as learning to walk, run and backflip from motion capture data; learning to navigate complex terrains; and by implementing a video game input control system. We show further robustness by training with alternative algorithms found in OpenAI.Baselines. Finally, we share strategies for significantly reducing the training time.