README.md

INENet: Inliers Estimation Network with Similarity Learning for Partial Overlapping Registration

by Yue Wu, Yue Zhang, Xiaolong Fan, Maoguo Gong, Qiguang Miao, Wenping Ma, and details are in paper.

Usage

1. Clone the repository.

2. Change the "DATA_DIR" parameter in line 14 of the "data_utils.py" file to its own data set folder path.

3. Run the "main.py" file.

Requirement

​ h5py=3.7.0

​ open3d=0.15.2

​ pytorch=1.11.0

​ pytorch3d=0.6.2

​ scikit-learn=1.1.1

​ transforms3d=0.4.1

​ tensorboardX=1.15.0

​ tqdm

​ numpy

Dataset

​ (1) ModelNet40

​ (2) 3DMatch

​ (3) S3DIS

Citation

If you find the code or trained models useful, please consider citing:

@article{2023inenet,
	author={Wu, Yue and Zhang, Yue and Fan, Xiaolong and Gong, Maoguo and Miao, Qiguang and Ma, Wenping},
	journal={IEEE Transactions on Circuits and Systems for Video Technology}, 
	title={INENet: Inliers Estimation Network With Similarity Learning for Partial Overlapping Registration}, 
	year={2023},
	volume={33},
	number={3},
	pages={1413-1426}
}

Acknowledgement

Our code refers to PointNet, DCP and MaskNet. We want to thank the above open-source projects.