https://github.com/NVlabs/Hydra-MDP

https://github.com/NVlabs/Hydra-MDP

Last synced: 7 days ago
JSON representation

• Host: GitHub
• URL: https://github.com/NVlabs/Hydra-MDP
• Owner: NVlabs
• Created: 2024-06-18T17:50:17.000Z (5 months ago)
• Default Branch: main
• Last Pushed: 2024-07-18T05:10:59.000Z (4 months ago)
• Last Synced: 2024-07-18T07:11:38.999Z (4 months ago)
• Size: 1.11 MB
• Stars: 151
• Watchers: 22
• Forks: 5
• Open Issues: 6
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Hydra-MDP: End-to-end Multimodal Planning with Multi-target Hydra-Distillation

![](./asset/hydra-mdp.png)

### [arXiv](https://arxiv.org/abs/2406.06978) | [Talk (EN)](https://opendrivelab.com/cvpr2024/workshop/) | [Talk (CN)](https://www.bilibili.com/video/BV1Pi421i7Ch/) | [Video](https://youtu.be/wfpLLSz5iWY?si=rVrKsO3oITTV-i1I) | [DriveLabs](https://www.youtube.com/watch?v=06BXs-R-fQ8) | [Blog](https://blogs.nvidia.com/blog/auto-research-cvpr-2024/) | [Challenge](https://opendrivelab.com/challenge2024/#end_to_end_driving_at_scale)

This repo contains the official implementation of Hydra-MDP. Hydra-MDP is a Transformer-based E2E Planning framework that uses Hydra-distillation to enable multi-target distillation from both human and rule-based teachers. 

## News

🏆Hydra-MDP won the 1st Place and Innovation Award of the [End-to-end Driving at Scale](https://opendrivelab.com/challenge2024/#end_to_end_driving_at_scale) Track at [CVPR24 Autonomous Grand Challenge](https://opendrivelab.com/challenge2024/).

## TODO

Delay in code and model release due to company policy. Stay tuned for updates.

## Citation

```

@article{li2024hydra,

  title={Hydra-MDP: End-to-end Multimodal Planning with Multi-target Hydra-Distillation},

  author={Li, Zhenxin and Li, Kailin and Wang, Shihao and Lan, Shiyi and Yu, Zhiding and Ji, Yishen and Li, Zhiqi and Zhu, Ziyue and Kautz, Jan and Wu, Zuxuan and others},

  journal={arXiv preprint arXiv:2406.06978},

  year={2024}

}

```

## Acknowledgements

Many thanks to the following great open-source repositories:

+ [NAVSIM](https://github.com/autonomousvision/navsim)

+ [The VAD Series](https://github.com/hustvl/VAD)

+ [Transfuser](https://github.com/autonomousvision/transfuser)

+ [tuplan_garage](https://github.com/autonomousvision/tuplan_garage)