End-to-End Autonomous Driving through V2X Cooperation

arXiv | Demo | Code

Table of Contents:

1. Highlights
2. News
3. Getting Started
   • Installation
   • Prepare Dataset
   • Train/Eval
4. TODO List
5. License
6. Citation

Highlights

• UniV2X is the first cooperative autonomous driving framework that seamlessly integrates all key driving modules across diverse driving views into a unified network.

News

• 2025/01/26 Code & model initial release v1.0.
• 2024/12/10 UniV2X is accepted by AAAI 2025.
• 2024/04/02 UniV2X paper is available on arXiv.

Getting Started

• Installation
• Prepare Dataset
• Evaluation Example
• Train/Eval

TODO List

• Base-model code release (on DAIR-V2X)
• Base-model configs & checkpoints (on DAIR-V2X)
• Benchmark code release (on DAIR-V2X)
• Benchmark configs & checkpoints (on DAIR-V2X)
• Support more datasets

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{yu2024_univ2x,
 title={End-to-End Autonomous Driving through V2X Cooperation}, 
 author={Haibao Yu and Wenxian Yang and Jiaru Zhong and Zhenwei Yang and Siqi Fan and Ping Luo and Zaiqing Nie},
 booktitle={The 39th Annual AAAI Conference on Artificial Intelligence},
 year={2025}
}

Related resources

• DAIR-V2X (:rocket:Ours!)
• V2X-Seq (:rocket:Ours!)
• FFNET (:rocket:Ours!)
• UniAD