Vision 3D

A clean, easy-to-use PyTorch library for lidar perception. Currently supports SECOND detector.

Project goals

• Emphasis on simple codebase (no 1,000 LOC functions).
• General 3D detection library (easy to extend to new models and datasets).
• Hope to reproduce state-of-the-art results.

Status and plans

• At this time I do not have capacity to develop this project. Community support is welcomed.
• I hope this project can serve as useful starting point for lidar perception research.
• Implementation of PV-RCNN is work-in-progress.
• These forks (one, two) have shown some promise in training on other datasets (NuScenes, and proprietary lidar data).

Usage

See inference.py and train.py. To train, need to first start a visdom server using command

visdom

Installation

See install.md.

Preliminary results

Citing

If you find this work helpful in your research, please consider starring this repo and citing:

@article{pvrcnnpytorch,
  author={Jacob Hultman},
  title={vision3d},
  journal={https://github.com/jhultman/vision3d},
  year={2020}
}

Contributions

Contributions are welcome. Please post an issue if you find any bugs.

Acknowledgements and licensing

Please see license.md. Note that the code in

vision3d/ops