Pc-net: Unsupervised point correspondence learning with neural networks

Introduction

This is the source code for our paper Pc-net: Unsupervised point correspondence learning with neural networks

Network Architecture

The architecture of our proposed model is as follows

---

Illustration of shape morphing.

Installation

• Our code is tested on Tensorflow>1.4 and CUDA 8.0

• Clone this repo

git clone https://github.com/lixiang-ucas/PCNet.git

Dataset

• Download the dataset from

Training

run the notebook: [Train]Fish_Def=0.4.ipynb

Test

run the notebook: [Test]Fish_Def=0.4.ipynb

Citation

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

@inproceedings{li2019pc,
  title={Pc-net: Unsupervised point correspondence learning with neural networks},
  author={Li, Xiang and Wang, Lingjing and Fang, Yi},
  booktitle={2019 International Conference on 3D Vision (3DV)},
  pages={145--154},
  year={2019},
  organization={IEEE}
}