Unlocking Generalization Power in LiDAR Point Cloud Registration

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zeng, Zhenxuan, Wu, Qiao, Zhang, Xiyu, Wu, Lin Yuanbo, An, Pei, Yang, Jiaqi, Wang, Ji, Wang, Peng
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866910872933236736
author Zeng, Zhenxuan
Wu, Qiao
Zhang, Xiyu
Wu, Lin Yuanbo
An, Pei
Yang, Jiaqi
Wang, Ji
Wang, Peng
author_facet Zeng, Zhenxuan
Wu, Qiao
Zhang, Xiyu
Wu, Lin Yuanbo
An, Pei
Yang, Jiaqi
Wang, Ji
Wang, Peng
contents In real-world environments, a LiDAR point cloud registration method with robust generalization capabilities (across varying distances and datasets) is crucial for ensuring safety in autonomous driving and other LiDAR-based applications. However, current methods fall short in achieving this level of generalization. To address these limitations, we propose UGP, a pruned framework designed to enhance generalization power for LiDAR point cloud registration. The core insight in UGP is the elimination of cross-attention mechanisms to improve generalization, allowing the network to concentrate on intra-frame feature extraction. Additionally, we introduce a progressive self-attention module to reduce ambiguity in large-scale scenes and integrate Bird's Eye View (BEV) features to incorporate semantic information about scene elements. Together, these enhancements significantly boost the network's generalization performance. We validated our approach through various generalization experiments in multiple outdoor scenes. In cross-distance generalization experiments on KITTI and nuScenes, UGP achieved state-of-the-art mean Registration Recall rates of 94.5% and 91.4%, respectively. In cross-dataset generalization from nuScenes to KITTI, UGP achieved a state-of-the-art mean Registration Recall of 90.9%. Code will be available at https://github.com/peakpang/UGP.
format Preprint
id arxiv_https___arxiv_org_abs_2503_10149
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Unlocking Generalization Power in LiDAR Point Cloud Registration
Zeng, Zhenxuan
Wu, Qiao
Zhang, Xiyu
Wu, Lin Yuanbo
An, Pei
Yang, Jiaqi
Wang, Ji
Wang, Peng
Computer Vision and Pattern Recognition
In real-world environments, a LiDAR point cloud registration method with robust generalization capabilities (across varying distances and datasets) is crucial for ensuring safety in autonomous driving and other LiDAR-based applications. However, current methods fall short in achieving this level of generalization. To address these limitations, we propose UGP, a pruned framework designed to enhance generalization power for LiDAR point cloud registration. The core insight in UGP is the elimination of cross-attention mechanisms to improve generalization, allowing the network to concentrate on intra-frame feature extraction. Additionally, we introduce a progressive self-attention module to reduce ambiguity in large-scale scenes and integrate Bird's Eye View (BEV) features to incorporate semantic information about scene elements. Together, these enhancements significantly boost the network's generalization performance. We validated our approach through various generalization experiments in multiple outdoor scenes. In cross-distance generalization experiments on KITTI and nuScenes, UGP achieved state-of-the-art mean Registration Recall rates of 94.5% and 91.4%, respectively. In cross-dataset generalization from nuScenes to KITTI, UGP achieved a state-of-the-art mean Registration Recall of 90.9%. Code will be available at https://github.com/peakpang/UGP.
title Unlocking Generalization Power in LiDAR Point Cloud Registration
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.10149