{"ID":2834592,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2512.01850","arxiv_id":"2512.01850","title":"Register Any Point: Scaling 3D Point Cloud Registration by Flow Matching","abstract":"Point cloud registration aligns multiple unposed point clouds into a common reference frame and is a core step for 3D reconstruction and robot localization without initial guess. In this work, we cast registration as conditional generation: a learned, continuous point-wise velocity field transports noisy points to a registered scene, from which the pose of each view is recovered. Unlike prior methods that perform correspondence matching to estimate pairwise transformations and then optimize a pose graph for multi-view registration, our model directly generates the registered point cloud, yielding both efficiency and point-level global consistency. By scaling the training data and conducting test-time rigidity enforcement, our approach achieves state-of-the-art results on existing pairwise registration benchmarks and on our proposed cross-domain multi-view registration benchmark. The superior zero-shot performance on this benchmark shows that our method generalizes across view counts, scene scales, and sensor modalities even with low overlap. Source code available at: https://github.com/PRBonn/RAP.","short_abstract":"Point cloud registration aligns multiple unposed point clouds into a common reference frame and is a core step for 3D reconstruction and robot localization without initial guess. In this work, we cast registration as conditional generation: a learned, continuous point-wise velocity field transports noisy points to a re...","url_abs":"https://arxiv.org/abs/2512.01850","url_pdf":"https://arxiv.org/pdf/2512.01850v2","authors":"[\"Yue Pan\",\"Tao Sun\",\"Liyuan Zhu\",\"Lucas Nunes\",\"Iro Armeni\",\"Jens Behley\",\"Cyrill Stachniss\"]","published":"2025-12-01T16:36:51Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.RO\"]","methods":"[]","has_code":false,"code_links":[{"ID":606426,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2834592,"paper_url":"https://arxiv.org/abs/2512.01850","paper_title":"Register Any Point: Scaling 3D Point Cloud Registration by Flow Matching","repo_url":"https://github.com/PRBonn/RAP","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}