TAU-R1: Visual Language Model for Traffic Anomaly Understanding

TAU-R1(Traffic Anomaly Understanding - R1) is a VLM-based two-layer hiearchichal framework aiming to solve traffic anomaly understanding task on real-world roundabout scenarios.

Dataset

The dataset for this project, Roundabout-TAU, is available on Hugging Face.

Code

Code for this project will be released later.

Citation

If you find this project helpful, please cite:

@misc{lin2026taur1visuallanguagemodel,
  title        = {TAU-R1: Visual Language Model for Traffic Anomaly Understanding},
  author       = {Yuqiang Lin and Kehua Chen and Sam Lockyer and Arjun Yadav and Mingxuan Sui and Shucheng Zhang and Yan Shi and Bingzhang Wang and Yuang Zhang and Markus Zarbock and Florain Stanek and Adrian Evans and Wenbin Li and Yinhai Wang and Nic Zhang},
  year         = {2026},
  eprint       = {2603.19098},
  archivePrefix= {arXiv},
  primaryClass = {cs.CV},
  url          = {https://arxiv.org/abs/2603.19098}
}

License

This project is licensed under the MIT License - see the LICENSE file for details.