haooozi / P2P

A Strong Tracking Framework for 3D SOT on LiDAR Point Clouds

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

P2P

The official implementation of the paper:

P2P: Part-to-Part Motion Cues Guide a Strong Tracking Framework for LiDAR Point Clouds

Jiahao Nie, Fei Xie, Sifan Zhou, Xueyi Zhou, Dong-Kyu Chae, Zhiwei He.

📜 [technical report], 🤗 [model weights]

🔥 Highlights

P2P is a strong tracking framework for 3D SOT on LiDAR point clouds that have:

  • Elegant tracking pipeline.
  • SOTA performance on KITTI, NuScenes and Waymo.
  • High efficiency.

📢 News

Important

If you have any question for our codes or model weights, please feel free to concat me at jhnie@hdu.edu.cn.

  • [2024/07/10] We released the arxiv version at here.
  • [2024/07/07] We released the installation, training, and testing details.
  • [2024/07/06] We released the implementation of our model.

📋 TODO List

  • All caterogy model weights of point and voxel versions trained on KITTI, Nuscenes.

🕹️ Getting Started

❤️ Acknowledgement

Our implementation is based on Open3DSOT, BEVTrack and MMDetection3D. Thanks for the great open-source work!

⭐ Citation

If any parts of our paper and code help your research, please consider citing us and giving a star to our repository.

@article{p2p,
 title={P2P: Part-to-Part Motion Cues Guide a Strong Tracking Framework for LiDAR Point Clouds},
 year={2024}
}

About

A Strong Tracking Framework for 3D SOT on LiDAR Point Clouds

License:MIT License


Languages

Language:Python 99.4%Language:Shell 0.6%