MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learning

Quanyi Li1,2*,  Zhenghao Peng1*,  Zhenghai Xue1Qihang Zhang1Bolei Zhou1 
1The Chinese University of Hong Kong, 2Centre for Perceptual and Interactive Intelligence
Webpage | Code | Paper
MetaDrive Simulator

To facilitate the research of generalizable reinforcement learning, we develop an open-source, highly efficient and flexible driving simulator MetaDrive, which holds the following key features:

We construct a variety of RL tasks and baselines in both single-agent and multi-agent settings, including benchmarking generalizability across unseen scenes, safe exploration, and learning multi-agent traffic.

Reference
Please refer to the technical report or the code repo for more information.
@misc{li2021metadrive,
      title={MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learning},
      author={Quanyi Li and Zhenghao Peng and Zhenghai Xue and Qihang Zhang and Bolei Zhou},
      year={2021},
      eprint={2109.12674},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}