PCRDiffusion: Diffusion Probabilistic Models for Point Cloud Registration

We propose a new framework that formulates point cloud registration as a denoising diffusion process from noisy transformation to object transformation. During training stage, object transformation diffuses from ground-truth transformation to random distribution, and the model learns to reverse this...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wu, Yue, Yuan, Yongzhe, Fan, Xiaolong, Huang, Xiaoshui, Gong, Maoguo, Miao, Qiguang
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Wu, Yue
Yuan, Yongzhe
Fan, Xiaolong
Huang, Xiaoshui
Gong, Maoguo
Miao, Qiguang
description We propose a new framework that formulates point cloud registration as a denoising diffusion process from noisy transformation to object transformation. During training stage, object transformation diffuses from ground-truth transformation to random distribution, and the model learns to reverse this noising process. In sampling stage, the model refines randomly generated transformation to the output result in a progressive way. We derive the variational bound in closed form for training and provide implementations of the model. Our work provides the following crucial findings: (i) In contrast to most existing methods, our framework, Diffusion Probabilistic Models for Point Cloud Registration (PCRDiffusion) does not require repeatedly update source point cloud to refine the predicted transformation. (ii) Point cloud registration, one of the representative discriminative tasks, can be solved by a generative way and the unified probabilistic formulation. Finally, we discuss and provide an outlook on the application of diffusion model in different scenarios for point cloud registration. Experimental results demonstrate that our model achieves competitive performance in point cloud registration. In correspondence-free and correspondence-based scenarios, PCRDifussion can both achieve exceeding 50\% performance improvements.
doi_str_mv 10.48550/arxiv.2312.06063
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2312_06063</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2312_06063</sourcerecordid><originalsourceid>FETCH-LOGICAL-a673-67f9d31587f68b78ee68c6ac37de562f9fa4d77bf4da7e39a7aa257d875f5a033</originalsourceid><addsrcrecordid>eNo9j81OwzAQhH3hgFoegBN-gQQnG3ud3lD4q1REVPUebWIvshRq5KQI3p5SUE8z0ugb6RPiulB5ZbVWt5S-wmdeQlHmyigDl2LdNtv7wHyYQtyv5LnKNsWe-jCGaQ6DfInOj5PkmGQbw36WzRgPTm7923FPNB-JpbhgGid_9Z8LsXt82DXP2eb1ad3cbTIyCJlBrh0U2iIb26P13tjB0ADovDYl10yVQ-y5coQeakKiUqOzqFmTAliIm7_bk0v3kcI7pe_u16k7OcEPCb1HjA</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>PCRDiffusion: Diffusion Probabilistic Models for Point Cloud Registration</title><source>arXiv.org</source><creator>Wu, Yue ; Yuan, Yongzhe ; Fan, Xiaolong ; Huang, Xiaoshui ; Gong, Maoguo ; Miao, Qiguang</creator><creatorcontrib>Wu, Yue ; Yuan, Yongzhe ; Fan, Xiaolong ; Huang, Xiaoshui ; Gong, Maoguo ; Miao, Qiguang</creatorcontrib><description>We propose a new framework that formulates point cloud registration as a denoising diffusion process from noisy transformation to object transformation. During training stage, object transformation diffuses from ground-truth transformation to random distribution, and the model learns to reverse this noising process. In sampling stage, the model refines randomly generated transformation to the output result in a progressive way. We derive the variational bound in closed form for training and provide implementations of the model. Our work provides the following crucial findings: (i) In contrast to most existing methods, our framework, Diffusion Probabilistic Models for Point Cloud Registration (PCRDiffusion) does not require repeatedly update source point cloud to refine the predicted transformation. (ii) Point cloud registration, one of the representative discriminative tasks, can be solved by a generative way and the unified probabilistic formulation. Finally, we discuss and provide an outlook on the application of diffusion model in different scenarios for point cloud registration. Experimental results demonstrate that our model achieves competitive performance in point cloud registration. In correspondence-free and correspondence-based scenarios, PCRDifussion can both achieve exceeding 50\% performance improvements.</description><identifier>DOI: 10.48550/arxiv.2312.06063</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Computer Vision and Pattern Recognition</subject><creationdate>2023-12</creationdate><rights>http://creativecommons.org/licenses/by/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2312.06063$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2312.06063$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Wu, Yue</creatorcontrib><creatorcontrib>Yuan, Yongzhe</creatorcontrib><creatorcontrib>Fan, Xiaolong</creatorcontrib><creatorcontrib>Huang, Xiaoshui</creatorcontrib><creatorcontrib>Gong, Maoguo</creatorcontrib><creatorcontrib>Miao, Qiguang</creatorcontrib><title>PCRDiffusion: Diffusion Probabilistic Models for Point Cloud Registration</title><description>We propose a new framework that formulates point cloud registration as a denoising diffusion process from noisy transformation to object transformation. During training stage, object transformation diffuses from ground-truth transformation to random distribution, and the model learns to reverse this noising process. In sampling stage, the model refines randomly generated transformation to the output result in a progressive way. We derive the variational bound in closed form for training and provide implementations of the model. Our work provides the following crucial findings: (i) In contrast to most existing methods, our framework, Diffusion Probabilistic Models for Point Cloud Registration (PCRDiffusion) does not require repeatedly update source point cloud to refine the predicted transformation. (ii) Point cloud registration, one of the representative discriminative tasks, can be solved by a generative way and the unified probabilistic formulation. Finally, we discuss and provide an outlook on the application of diffusion model in different scenarios for point cloud registration. Experimental results demonstrate that our model achieves competitive performance in point cloud registration. In correspondence-free and correspondence-based scenarios, PCRDifussion can both achieve exceeding 50\% performance improvements.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Computer Vision and Pattern Recognition</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNo9j81OwzAQhH3hgFoegBN-gQQnG3ud3lD4q1REVPUebWIvshRq5KQI3p5SUE8z0ugb6RPiulB5ZbVWt5S-wmdeQlHmyigDl2LdNtv7wHyYQtyv5LnKNsWe-jCGaQ6DfInOj5PkmGQbw36WzRgPTm7923FPNB-JpbhgGid_9Z8LsXt82DXP2eb1ad3cbTIyCJlBrh0U2iIb26P13tjB0ADovDYl10yVQ-y5coQeakKiUqOzqFmTAliIm7_bk0v3kcI7pe_u16k7OcEPCb1HjA</recordid><startdate>20231210</startdate><enddate>20231210</enddate><creator>Wu, Yue</creator><creator>Yuan, Yongzhe</creator><creator>Fan, Xiaolong</creator><creator>Huang, Xiaoshui</creator><creator>Gong, Maoguo</creator><creator>Miao, Qiguang</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20231210</creationdate><title>PCRDiffusion: Diffusion Probabilistic Models for Point Cloud Registration</title><author>Wu, Yue ; Yuan, Yongzhe ; Fan, Xiaolong ; Huang, Xiaoshui ; Gong, Maoguo ; Miao, Qiguang</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a673-67f9d31587f68b78ee68c6ac37de562f9fa4d77bf4da7e39a7aa257d875f5a033</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Computer Vision and Pattern Recognition</topic><toplevel>online_resources</toplevel><creatorcontrib>Wu, Yue</creatorcontrib><creatorcontrib>Yuan, Yongzhe</creatorcontrib><creatorcontrib>Fan, Xiaolong</creatorcontrib><creatorcontrib>Huang, Xiaoshui</creatorcontrib><creatorcontrib>Gong, Maoguo</creatorcontrib><creatorcontrib>Miao, Qiguang</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Wu, Yue</au><au>Yuan, Yongzhe</au><au>Fan, Xiaolong</au><au>Huang, Xiaoshui</au><au>Gong, Maoguo</au><au>Miao, Qiguang</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>PCRDiffusion: Diffusion Probabilistic Models for Point Cloud Registration</atitle><date>2023-12-10</date><risdate>2023</risdate><abstract>We propose a new framework that formulates point cloud registration as a denoising diffusion process from noisy transformation to object transformation. During training stage, object transformation diffuses from ground-truth transformation to random distribution, and the model learns to reverse this noising process. In sampling stage, the model refines randomly generated transformation to the output result in a progressive way. We derive the variational bound in closed form for training and provide implementations of the model. Our work provides the following crucial findings: (i) In contrast to most existing methods, our framework, Diffusion Probabilistic Models for Point Cloud Registration (PCRDiffusion) does not require repeatedly update source point cloud to refine the predicted transformation. (ii) Point cloud registration, one of the representative discriminative tasks, can be solved by a generative way and the unified probabilistic formulation. Finally, we discuss and provide an outlook on the application of diffusion model in different scenarios for point cloud registration. Experimental results demonstrate that our model achieves competitive performance in point cloud registration. In correspondence-free and correspondence-based scenarios, PCRDifussion can both achieve exceeding 50\% performance improvements.</abstract><doi>10.48550/arxiv.2312.06063</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2312.06063
ispartof
issn
language eng
recordid cdi_arxiv_primary_2312_06063
source arXiv.org
subjects Computer Science - Artificial Intelligence
Computer Science - Computer Vision and Pattern Recognition
title PCRDiffusion: Diffusion Probabilistic Models for Point Cloud Registration
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-19T18%3A32%3A17IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=PCRDiffusion:%20Diffusion%20Probabilistic%20Models%20for%20Point%20Cloud%20Registration&rft.au=Wu,%20Yue&rft.date=2023-12-10&rft_id=info:doi/10.48550/arxiv.2312.06063&rft_dat=%3Carxiv_GOX%3E2312_06063%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true