CurrI2P: Inter- and Intra-Modality Similarity Curriculum Learning for Image-to-Point Cloud Registration

CurrI2P: Inter- and Intra-Modality Similarity Curriculum Learning for Image-to-Point Cloud Registration | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article CurrI2P: Inter- and Intra-Modality Similarity Curriculum Learning for Image-to-Point Cloud Registration Liwei Lin, Chunyu Lin, Lang Nie, Shujuan Huang, Yao Zhao This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5457151/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Apr, 2025 Read the published version in The Visual Computer → Version 1 posted 11 You are reading this latest preprint version Abstract Image-to-point cloud (I2P) registration aims to align 2D images and 3D point clouds within the same scene. However, existing methods neglect the differences in learning difficulties among different samples, failing to bridge the image-to-point cloud modality gap adequately. Moreover, the features of images or point clouds within the same modality are easily affected by various geometric transformations, which further increases the complexity of the registration task. To address these issues, we propose a similarity curriculum learning framework for I2P registration, termed CurrI2P, which not only effectively captures similarities between different data pairs but also gradually adapts the model to more complex data transformations. Specifically, we first present inter-modality similarity curriculum learning to progressively enhance cross-modality similarity to reduce inaccurate matches caused by coarse registration at the early training stage. In addition, we design intra-modality similarity curriculum learning, particularly targeting to learn invariant features of images or point clouds within the same modality under the geometric disturbance of increasing magnitude. In experiments, we incorporate CurrI2P into different baselines to validate the feasibility and achieve state-of-the-art (SoTA) performance on both KITTI and nuScenes datasets. The code is available at https://github.com/lin-liwei/CurrI2P . Image-to-point cloud registration Similarity curriculum learning Inter-modality similarity Intra-modality similarity Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 07 Apr, 2025 Read the published version in The Visual Computer → Version 1 posted Editorial decision: Revision requested 22 Feb, 2025 Reviews received at journal 20 Feb, 2025 Reviewers agreed at journal 17 Feb, 2025 Reviewers agreed at journal 05 Dec, 2024 Reviewers agreed at journal 04 Dec, 2024 Reviews received at journal 24 Nov, 2024 Reviewers agreed at journal 24 Nov, 2024 Reviewers invited by journal 22 Nov, 2024 Editor assigned by journal 15 Nov, 2024 Submission checks completed at journal 15 Nov, 2024 First submitted to journal 14 Nov, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5457151","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":387451005,"identity":"aea0d64a-339b-4534-abcb-7d6fc79bd6b9","order_by":0,"name":"Liwei Lin","email":"","orcid":"","institution":"Beijing Jiaotong University","correspondingAuthor":false,"prefix":"","firstName":"Liwei","middleName":"","lastName":"Lin","suffix":""},{"id":387451006,"identity":"4563a54a-3b79-4554-80c0-3b2b109b0a0c","order_by":1,"name":"Chunyu Lin","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAw0lEQVRIiWNgGAWjYFAC5oYDDBUHmMFsHuK0MAK1nDnAzEOSFgbGtgMMxGvhn5HYeLhw3h12e4kExgdv2xjkzQlpkbiR2HB45rZnzDwSCcyGc9sYDHc2ENBiIAHUwrvtMEgLmzRvG0OCwQGitMwBa2H/TYKWBogtzERpkTjzsOEwzzGgX848bJacc07CcAMhLfztyYc/89TcSWZvTz744U2ZjTxBWxgEEsBUMjiCgLYSUg+yBmKoHRFKR8EoGAWjYKQCAGB4P+zwj7aIAAAAAElFTkSuQmCC","orcid":"","institution":"Beijing Jiaotong University","correspondingAuthor":true,"prefix":"","firstName":"Chunyu","middleName":"","lastName":"Lin","suffix":""},{"id":387451008,"identity":"b6540263-31e8-4789-8137-0db30bc923ff","order_by":2,"name":"Lang Nie","email":"","orcid":"","institution":"Beijing Jiaotong University","correspondingAuthor":false,"prefix":"","firstName":"Lang","middleName":"","lastName":"Nie","suffix":""},{"id":387451009,"identity":"7f1b99c7-d52b-47f0-b238-cfaa8f14ac8c","order_by":3,"name":"Shujuan Huang","email":"","orcid":"","institution":"Beijing Jiaotong University","correspondingAuthor":false,"prefix":"","firstName":"Shujuan","middleName":"","lastName":"Huang","suffix":""},{"id":387451010,"identity":"2f5b5ba0-b7a7-4c9e-9b1f-55f8814e55f5","order_by":4,"name":"Yao Zhao","email":"","orcid":"","institution":"Beijing Jiaotong University","correspondingAuthor":false,"prefix":"","firstName":"Yao","middleName":"","lastName":"Zhao","suffix":""}],"badges":[],"createdAt":"2024-11-15 02:53:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5457151/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5457151/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00371-025-03889-7","type":"published","date":"2025-04-07T16:05:36+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":80558937,"identity":"f0e5e5b0-b46e-447b-a683-dcd80de9f852","added_by":"auto","created_at":"2025-04-14 16:17:08","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":11103484,"visible":true,"origin":"","legend":"","description":"","filename":"CurrI2P.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5457151/v1_covered_81579e42-ee89-4c8b-a2dd-88b33e85f63d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"CurrI2P: Inter- and Intra-Modality Similarity Curriculum Learning for Image-to-Point Cloud Registration","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"the-visual-computer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"tvcj","sideBox":"Learn more about [The Visual Computer](http://link.springer.com/journal/371)","snPcode":"371","submissionUrl":"https://submission.nature.com/new-submission/371/3","title":"The Visual Computer","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Image-to-point cloud registration, Similarity curriculum learning, Inter-modality similarity, Intra-modality similarity","lastPublishedDoi":"10.21203/rs.3.rs-5457151/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5457151/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Image-to-point cloud (I2P) registration aims to align 2D images and 3D point clouds within the same scene. However, existing methods neglect the differences in learning difficulties among different samples, failing to bridge the image-to-point cloud modality gap adequately. Moreover, the features of images or point clouds within the same modality are easily affected by various geometric transformations, which further increases the complexity of the registration task. To address these issues, we propose a similarity curriculum learning framework for I2P registration, termed CurrI2P, which not only effectively captures similarities between different data pairs but also gradually adapts the model to more complex data transformations. Specifically, we first present inter-modality similarity curriculum learning to progressively enhance cross-modality similarity to reduce inaccurate matches caused by coarse registration at the early training stage. In addition, we design intra-modality similarity curriculum learning, particularly targeting to learn invariant features of images or point clouds within the same modality under the geometric disturbance of increasing magnitude. In experiments, we incorporate CurrI2P into different baselines to validate the feasibility and achieve state-of-the-art (SoTA) performance on both KITTI and nuScenes datasets. The code is available at https://github.com/lin-liwei/CurrI2P.","manuscriptTitle":"CurrI2P: Inter- and Intra-Modality Similarity Curriculum Learning for Image-to-Point Cloud Registration","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-13 06:09:08","doi":"10.21203/rs.3.rs-5457151/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-02-22T19:57:54+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-21T03:36:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"228054098633112854327502435484825451022","date":"2025-02-17T06:23:01+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"177701415140579356112557708220115246688","date":"2024-12-05T05:24:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"12114338084589526454481513684247220850","date":"2024-12-05T03:16:15+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-24T09:32:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"60984619907770129080890657503528772613","date":"2024-11-24T06:06:44+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-11-22T11:39:49+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-11-15T08:14:08+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-11-15T07:25:03+00:00","index":"","fulltext":""},{"type":"submitted","content":"The Visual Computer","date":"2024-11-15T02:50:11+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"the-visual-computer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"tvcj","sideBox":"Learn more about [The Visual Computer](http://link.springer.com/journal/371)","snPcode":"371","submissionUrl":"https://submission.nature.com/new-submission/371/3","title":"The Visual Computer","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"e721f1eb-2cfd-4ed4-abb8-74a7e30cc599","owner":[],"postedDate":"December 13th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-04-14T16:13:26+00:00","versionOfRecord":{"articleIdentity":"rs-5457151","link":"https://doi.org/10.1007/s00371-025-03889-7","journal":{"identity":"the-visual-computer","isVorOnly":false,"title":"The Visual Computer"},"publishedOn":"2025-04-07 16:05:36","publishedOnDateReadable":"April 7th, 2025"},"versionCreatedAt":"2024-12-13 06:09:08","video":"","vorDoi":"10.1007/s00371-025-03889-7","vorDoiUrl":"https://doi.org/10.1007/s00371-025-03889-7","workflowStages":[]},"version":"v1","identity":"rs-5457151","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5457151","identity":"rs-5457151","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}