HybridReg: Robust 3D Point Cloud Registration with Hybrid Motions

TL;DR

HybridReg addresses 3D point cloud registration in scenes with hybrid motions, where backgrounds are rigid while foregrounds deform. It introduces an uncertainty mask to separate motion types and a negative log-likelihood loss to model uncertainty during feature extraction and correlation. The authors also present HybridMatch, a scene-level dataset with controllable deforming foregrounds to benchmark the approach. Across HybridMatch, 3DMatch/3DLoMatch, and ETH, HybridReg achieves state-of-the-art performance and demonstrates robustness to varying non-rigid content, advancing practical registration in dynamic environments.

Abstract

Scene-level point cloud registration is very challenging when considering dynamic foregrounds. Existing indoor datasets mostly assume rigid motions, so the trained models cannot robustly handle scenes with non-rigid motions. On the other hand, non-rigid datasets are mainly object-level, so the trained models cannot generalize well to complex scenes. This paper presents HybridReg, a new approach to 3D point cloud registration, learning uncertainty mask to account for hybrid motions: rigid for backgrounds and non-rigid/rigid for instance-level foregrounds. First, we build a scene-level 3D registration dataset, namely HybridMatch, designed specifically with strategies to arrange diverse deforming foregrounds in a controllable manner. Second, we account for different motion types and formulate a mask-learning module to alleviate the interference of deforming outliers. Third, we exploit a simple yet effective negative log-likelihood loss to adopt uncertainty to guide the feature extraction and correlation computation. To our best knowledge, HybridReg is the first work that exploits hybrid motions for robust point cloud registration. Extensive experiments show HybridReg's strengths, leading it to achieve state-of-the-art performance on both widely-used indoor and outdoor datasets.