Home

Awesome

Contrastive Boundary Learning for Point Cloud Segmentation (CVPR 2022)

By Liyao Tang, Yibing Zhan, Zhe Chen, Baosheng Yu, and Dacheng Tao

This is the implementation of our CVPR 2022 paper: <br> Contrastive Boundary Learning for Point Cloud Segmentation [arXiv]

cbl

If you find our work useful in your research, please consider citing:

@InProceedings{tang2022cbl,
  author={Tang, Liyao and Zhan, Yibing and Chen, Zhe and Yu, Baosheng and Tao, Dacheng},
  booktitle={2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, 
  title={Contrastive Boundary Learning for Point Cloud Segmentation}, 
  year={2022},
  volume={},
  number={},
  pages={8479-8489},
  doi={10.1109/CVPR52688.2022.00830}
}

Setup & Usage

For point-transformer baseline, please follow pytorch/README.

For ConvNet and other baselines, please follow tensorflow/README.

Pre-trained models

Pretrained models can be accessed here, together with training and testing log. Choose the desired baseline and unzip into the corresponding code directory (tensorflow/pytorch) and follow the README there for further instruction.

Quantitative results

S3DIS (Area 5)

baselinemIoUOAmACC
ConvNet + CBL69.490.675.2
ConvNet + CBL (kl)69.590.975.3
point-transformer + CBL71.691.277.9

Qualitative results

demo

Acknowledgement

Codes are built based on a series of previous works, including: <br> KPConv, <br> RandLA-Net, <br> CloserLook3D, <br> Point-Transformer. <br> Thanks for their excellent work.

License

This repo is licensed under the terms of the MIT license (see LICENSE file for details).