NuSync

Project Page | Paper | GitHub

NuSync is a dataset for end-to-end autonomous driving, introduced as part of the paper AutoMoT: A Unified Vision-Language-Action Model with Asynchronous Mixture-of-Transformers for End-to-End Autonomous Driving.

Introduction

AutoMoT is an end-to-end autonomous driving (AD) framework that unifies reasoning and action generation within a single vision-language-action (VLA) model. It leverages an asynchronous Mixture-of-Transformers (MoT) architecture with joint attention sharing, which preserves the general reasoning capabilities of pre-trained VLMs while enabling efficient fast-slow inference through execution at different task frequencies.

The NuSync dataset is released alongside AutoMoT to support research in multi-task scene understanding and action-level tasks such as decision-making and trajectory planning in autonomous driving.

Citation

If you find this dataset or the AutoMoT model useful, please cite the following:

@article{huang2026automot,
  title   = {AutoMoT: A Unified Vision-Language-Action Model with Asynchronous Mixture-of-Transformers for End-to-End Autonomous Driving},
  author  = {Wenhui Huang and Songyan Zhang and Qihang Huang and Zhidong Wang and Zhiqi Mao and Collister Chua and Zhan Chen and Long Chen and Chen Lv},
  journal = {arXiv preprint arXiv:2603.14851},
  year    = {2026},
  url     = {https://arxiv.org/abs/2603.14851}
}

@inproceedings{jia2024bench,
  title     = {Bench2Drive: Towards Multi-Ability Benchmarking of Closed-Loop End-To-End Autonomous Driving},
  author    = {Xiaosong Jia and Zhenjie Yang and Qifeng Li and Zhiyuan Zhang and Junchi Yan},
  booktitle = {NeurIPS 2024 Datasets and Benchmarks Track},
  year      = {2024}
}