Skip to content

Latest commit

 

History

History

PUCRN

readme.md

Point Cloud Upsampling via Cascaded Refinement Network

Introduction

This is the official Pytorch implementation of our paper "Point Cloud Upsampling via Cascaded Refinement Network" (Paper) in Asian Conference on Computer Vision (ACCV) 2022.

Data Preparation

We use two public datasets, including PU1K (Paper, Download), and PUGAN (Paper, Download).

Usage

The code is tested under Pytorch 1.6.0 and Python 3.6 on Ubuntu 16.04. Pretrained weights are available in ./model/release.

  1. Install python denpendencies.
pip install -r requirements.txt
  1. Compile pyTorch extensions.
cd pointnet2_ops_lib
python setup.py install

cd ../losses
python setup.py install
  1. Train the model.
sh start_train.sh
  1. Evaluate the model.
cd evaluation_code
cmake .
make

sh test.sh

License

This project is released under the Apache 2.0 license. Other codes from open source repository follows the original distributive licenses.

Acknowledgement

We borrow some code from MPU, PU-Net, SnowflakeNet, and PUGCN.

Citation

Please consider citing our paper in your publications if the project helps your research. BibTeX reference is as follows.

@inproceedings{du2022point,
    title={Point Cloud Upsampling via Cascaded Refinement Network},
    author={Du, Hang and Yan, Xuejun and Wang, Jingjing and Xie, Di and Pu, Shiliang},
    booktitle={Asian Conference on Computer Vision},
    year={2022},
}