MDR-MFI:Multi-Branch Decoupled Regression and Multi-Scale Feature Interaction for Partial-to-Partial Cloud Registration
Weidong Dai, Xuejun Yan, Jingjing Wang, Di Xie, Shiliang Pu
Abstract
Point cloud registration is a fundamental task in the 3D vision field. Many previous works adopt the regression model to estimate the transformation parameters. However, these methods couple the estimation of rotation and translation via a single regression branch, which suffers from the mutual interference among rotation and translation. In addition, previous methods extract and interact features in a single scale, which ignores the rich information from multiple scales. To address above issues, in this paper, we propose a multi-branch decoupled regression and multi-scale feature interaction (MDR-MFI) framework for point cloud registration. Firstly, we decouple the estimation of 7 transformation parameters via multiple regression branches. The decoupled structure effectively mitigates the mutual interference among 7 parameters, resulting in improved performance. Secondly, we propose a multi-scale feature extraction and interaction framework to encourage the network to learn more discriminative features. Experimental results demonstrate that our method achieves state-of-the-art performance on public datasets. The code is available at https://github.com/hikvision-research/3DVision.
BibTeX
@inproceedings{icassp2023_mdrmfimultibranc,
title = {MDR-MFI:Multi-Branch Decoupled Regression and Multi-Scale Feature Interaction for Partial-to-Partial Cloud Registration},
author = {Weidong Dai and Xuejun Yan and Jingjing Wang and Di Xie and Shiliang Pu},
booktitle = {ICASSP 2023},
year = {2023}
}