GraVoS: Voxel Selection for 3D Point-Cloud Detection

This repository contains the PyTorch implementation of the CVPR'2023 paper - GraVoS: Voxel Selection for 3D Point-Cloud Detection.

Installation

The code was tested in the following environment:

• Ubuntu 18.04/20.04
• Python 3.7
• CUDA 11.1
• Pytorch 1.9.0
• spconv 2.1.21
• spconv-cu111 2.1.21

Create enviroment

conda create --name GraVoS python==3.7
conda activate GraVoS
pip install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html
git clone https://github.com/oshrout/GraVoS
cd GraVoS
pip install -r requirements.txt

Install spconv and OpenPCDet

1. Install spconv with pip, see spconv for more details.
2. Install pcdet by running python setup.py develop, see OpenPCDet for more details.

License

This project is realeased under the Apache License 2.0.

Citation

If you find this project useful, please consider cite:

@article{shrout2022gravos,
  title={GraVoS: Gradient based Voxel Selection for 3D Detection},
  author={Shrout, Oren and Ben-Shabat, Yizhak and Tal, Ayellet},
  journal={arXiv preprint arXiv:2208.08780},
  year={2022}
}

Acknowlegements

Our code is mostly based on OpenPCDet.