DriveE2E: Benchmarking Closed-Loop End-to-End Autonomous Driving Based-on Real-World Traffic Scenarios

arXiv | Demo | Code

Table of Contents:

1. Highlights
2. TODO List
3. License
4. Citation

Highlights

• DriveE2E is a simple yet challenging closed-loop evaluation framework, which closely integrates real-world driving scenarios into the CARLA simulator, effectively bridging the gap between simulated and real-world driving environments.

TODO List

• Closed-loop simulator release
• Base-model configs & checkpoints
• Benchmark code release
• Benchmark configs & checkpoints

License

All assets and code are under the Apache 2.0 license unless specified otherwise.

Citation

Please consider citing our paper if the project helps your research with the following BibTex:

@inproceedings{yu2025_drivee2e,
 title={DriveE2E: Benchmarking Closed-Loop End-to-End Autonomous Driving Based-on Real-World Traffic Scenarios}, 
 author={Haibao Yu and Wenxian Yang and Ruiyang Hao and Chuanye Wang and Jiaru Zhong and Ping Luo and Zaiqing Nie},
 booktitle={},
 year={2025}
}