Edge-Selective Feature Weaving for Point Cloud Matching
This repository contains a PyTorch-lightning implementation of the ESFW module proposed in our paper Edge-Selective Feature Weaving for Point Cloud Matching https://arxiv.org/pdf/2202.02149v1.pdf.
Note
Our code is created based on https://github.com/ZENGYIMING-EAMON/CorrNet3D
Installation
conda create --name corrnet3d python=3.8
conda activate corrnet3d
pip install pytorch-lightning==1.1.6
pip install --upgrade https://github.com/unlimblue/KNN_CUDA/releases/download/0.2/KNN_CUDA-0.2-py3-none-any.whl
pip install "git+git://github.com/erikwijmans/Pointnet2_PyTorch.git#egg=pointnet2_ops&subdirectory=pointnet2_ops_lib"
conda install torchvision torchaudio cudatoolkit=10.2 -c pytorch
pip install h5py
pip install tables
pip install matplotlib
pip install torch-scatter torch-sparse torch-cluster torch-spline-conv torch-geometric -f https://data.pyg.org whl/torch-1.10.0+cu102.html
Dockerfile
You can use my docker file
docker build ./ -t {image_name}
Datasets
Download from https://github.com/ZENGYIMING-EAMON/CorrNet3D
Train
uncomment 'cli_main()' in lit_corrnet3d_ESFW.py
python lit_corrnet3d_ESFW.py --batch_size=10 --data_dir=./trainset.h5 --test_data_dir=./testset.h5 --num_gpus
Test
To test on the whole testing set, run:
uncomment 'cli_main_test_()' in lit_corrnet3d_ESFW.py
python lit_corrnet3d_ESFW.py --batch_size=1 --ckpt_user=
--data_dir=./trainset.h5 --test_data_dir=./testset.h5 -- num_gpus
How to cite
@article{yanagi2022edge,
title={Edge-selective feature weaving for point cloud matching},
author={Zhou, Wengang and Li, Houqiang and Tian, Qi},
author={Yanagi, Rintaro and Atsushi, Hashimoto and Shusaku, Sone and Naoya, Chiba and Jiaxin, Ma and Yoshitaka, Ushiku},
journal={arXiv preprint arXiv:2202.02149},
year={2021}
}