OpenDriveLab / HDGT

[IEEE T-PAMI 2023] Unified heterogeneous transformer-based graph neural network for motion prediction

HDGT: Modeling the Driving Scene with Heterogenity and Relativity

HDGT: Heterogeneous Driving Graph Transformer for Multi-Agent Trajectory Prediction via Scene Encoding [IEEE TPAMI 2023]

• Paper

Introduction

HDGT is an unified heterogeneous transformer-based graph neural network for driving scene encoding. It is a SOTA method on INTERACTION and Waymo Motion Prediction Chanllege.

By time of release in April 2022, the proposed method achieves new state-of-the-art on INTERACTION Prediction Challenge and Waymo Open Motion Challenge, in which we rank the first and second respectively in terms of the minADE/minFDE metric.

Getting Started

• Installation
• Prepare Dataset
• Train & Evaluation

License

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

Bibtex

If this work is helpful for your research, please consider citing the following BibTeX entry.

@article{jia2023hdgt,
  title={HDGT: Heterogeneous Driving Graph Transformer for Multi-Agent Trajectory Prediction via Scene Encoding},
  author={Jia, Xiaosong and Wu, Penghao and Chen, Li and  Liu, Yu  and Li, Hongyang and Yan, Junchi},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)},
  year = {2023},
}  

@inproceedings{jia2022temporal,
  title={Towards Capturing the Temporal Dynamics for Trajectory Prediction: a Coarse-to-Fine Approach},
  author={Jia, Xiaosong and Chen, Li and Wu, Penghao and Zeng, Jia and  Yan, Junchi and Li, Hongyang and Qiao, Yu},
  booktitle={CoRL},
  year={2022}
}