SFL-GS: Spatio-Temporal Feature-guided Learning for 3D Gaussian Segmentation

SFL-GS: Spatio-Temporal Feature-guided Learning for 3D Gaussian Segmentation | 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 SFL-GS: Spatio-Temporal Feature-guided Learning for 3D Gaussian Segmentation Fang Wan, Zhiwei Ye, Xianjin Shi, Tianyu Li, Guangbo Lei, Li Xu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6664114/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 20 You are reading this latest preprint version Abstract 3D Gaussian splatting has emerged as a promising technique for real-time scene representation, making interactive 3D segmentation increasingly important in scene manipulation. However, inconsistent results generated by 2D segmentation across different viewpoints pose significant challenges for learning 3D segmentation feature fields. The accuracy of 3D segmentation decreases substantially when cross-view 2D segmentation results conflict. To address this issue, we present Spatio-Temporal Feature-guided Learning for 3D Gaussian Segmentation (SFL-GS), an efficient interactive 3D segmentation framework. SFL-GS employs a novel Spatio-temporal Feature-guided Learning (SFL) strategy to capture spatio-temporally consistent features and masks from 2D segmentation results across different views, thereby coherently guiding the learning of 3D segmentation feature fields. To refine feature and mask consistency in complex scenes, particularly under severe occlusion, our framework develops an enhanced optimization strategy that integrates statistical filtering, dynamic scale growth, and edge-aware optimization. This approach results in clearer boundaries and significantly improves segmentation accuracy, even in challenging environments. Extensive experiments demonstrate that our method achieves superior accuracy in segmentation tasks, making it suitable for precise and efficient 3D segmentation requirements in real-world applications. 3D Gaussian splatting Spatio-Temporal Feature-guided Learning 3D segmentation enhanced optimization strategy Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 30 Jul, 2025 Reviews received at journal 28 Jul, 2025 Reviews received at journal 27 Jul, 2025 Reviews received at journal 27 Jul, 2025 Reviews received at journal 24 Jul, 2025 Reviews received at journal 21 Jul, 2025 Reviewers agreed at journal 11 Jul, 2025 Reviewers agreed at journal 11 Jul, 2025 Reviewers agreed at journal 10 Jul, 2025 Reviewers agreed at journal 09 Jul, 2025 Reviews received at journal 09 Jul, 2025 Reviewers agreed at journal 09 Jul, 2025 Reviewers agreed at journal 09 Jul, 2025 Reviewers agreed at journal 08 Jul, 2025 Reviewers agreed at journal 08 Jul, 2025 Reviewers agreed at journal 08 Jul, 2025 Reviewers invited by journal 08 Jul, 2025 Editor assigned by journal 08 Jul, 2025 Submission checks completed at journal 20 May, 2025 First submitted to journal 14 May, 2025 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-6664114","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":483126867,"identity":"1d2379a2-420e-467b-b481-f9e497b09ef9","order_by":0,"name":"Fang Wan","email":"","orcid":"","institution":"Hubei University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Fang","middleName":"","lastName":"Wan","suffix":""},{"id":483126868,"identity":"dfbbb3e9-19b3-434e-b7e3-78726bbac433","order_by":1,"name":"Zhiwei Ye","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA10lEQVRIiWNgGAWjYBACAygtZyABppmJ12IM1MLYQJKWxA1EazHnX/z4M8+fw+nbpXvMHzBUWCc2sJ89gFeL5YxnZtI8PIdzd845Y9jAcCY9sYEnLwG/w26cYWPmkbidu+FGjmEDY9vhxAYJHgNCWpg/8xjcTjcAa/lHjJbzPQzSPAm3EyBaGoiyhc1Mcs6B/4Yb7hwrnJFwLN24jSeHkC2HH3948ydN3uB284YPH2qsZfvZz+DXwiCRwMDEA+MkADEbfvVAwH+AgfEHQVWjYBSMglEwogEAr8xKbDw77gkAAAAASUVORK5CYII=","orcid":"","institution":"Hubei University of Technology","correspondingAuthor":true,"prefix":"","firstName":"Zhiwei","middleName":"","lastName":"Ye","suffix":""},{"id":483126869,"identity":"dff277f7-d291-4521-a429-3ebd6f3cc97c","order_by":2,"name":"Xianjin Shi","email":"","orcid":"","institution":"Hubei University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Xianjin","middleName":"","lastName":"Shi","suffix":""},{"id":483126870,"identity":"4d7f636a-d333-46a2-a2a2-9a380b0a19ab","order_by":3,"name":"Tianyu Li","email":"","orcid":"","institution":"Peking University","correspondingAuthor":false,"prefix":"","firstName":"Tianyu","middleName":"","lastName":"Li","suffix":""},{"id":483126871,"identity":"f9c2f5ff-040d-4974-980e-9b9d691a3742","order_by":4,"name":"Guangbo Lei","email":"","orcid":"","institution":"Hubei University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Guangbo","middleName":"","lastName":"Lei","suffix":""},{"id":483126872,"identity":"b7d37eae-1488-421a-8df7-3d27204a69cf","order_by":5,"name":"Li Xu","email":"","orcid":"","institution":"Hubei University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Xu","suffix":""}],"badges":[],"createdAt":"2025-05-14 12:23:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6664114/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6664114/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":86473520,"identity":"3a787484-d082-4b76-a6d8-65a236ac011d","added_by":"auto","created_at":"2025-07-11 06:08:36","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1131230,"visible":true,"origin":"","legend":"","description":"","filename":"SFLGS.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6664114/v1_covered_97143ca4-672e-4bd4-b2ac-43cd77f44d43.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"SFL-GS: Spatio-Temporal Feature-guided Learning for 3D Gaussian Segmentation","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-journal-of-supercomputing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [The Journal of Supercomputing](https://www.springer.com/journal/11227)","snPcode":"11227","submissionUrl":"https://submission.nature.com/new-submission/11227/3","title":"The Journal of Supercomputing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"3D Gaussian splatting, Spatio-Temporal Feature-guided Learning, 3D segmentation, enhanced optimization strategy","lastPublishedDoi":"10.21203/rs.3.rs-6664114/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6664114/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e3D Gaussian splatting has emerged as a promising technique for real-time scene representation, making interactive 3D segmentation increasingly important in scene manipulation. However, inconsistent results generated by 2D segmentation across different viewpoints pose significant challenges for learning 3D segmentation feature fields. The accuracy of 3D segmentation decreases substantially when cross-view 2D segmentation results conflict. To address this issue, we present Spatio-Temporal Feature-guided Learning for 3D Gaussian Segmentation (SFL-GS), an efficient interactive 3D segmentation framework. SFL-GS employs a novel Spatio-temporal Feature-guided Learning (SFL) strategy to capture spatio-temporally consistent features and masks from 2D segmentation results across different views, thereby coherently guiding the learning of 3D segmentation feature fields. To refine feature and mask consistency in complex scenes, particularly under severe occlusion, our framework develops an enhanced optimization strategy that integrates statistical filtering, dynamic scale growth, and edge-aware optimization. This approach results in clearer boundaries and significantly improves segmentation accuracy, even in challenging environments. Extensive experiments demonstrate that our method achieves superior accuracy in segmentation tasks, making it suitable for precise and efficient 3D segmentation requirements in real-world applications.\u003c/p\u003e","manuscriptTitle":"SFL-GS: Spatio-Temporal Feature-guided Learning for 3D Gaussian Segmentation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-11 05:52:30","doi":"10.21203/rs.3.rs-6664114/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-30T20:17:47+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-29T03:39:00+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-27T22:28:54+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-27T13:33:13+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-24T07:45:59+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-21T07:22:53+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"105731553228516902692884947193737952872","date":"2025-07-12T02:03:45+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"220378634396068307061276981050755140418","date":"2025-07-11T09:28:55+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"245028136762847966105314745813318923184","date":"2025-07-10T09:20:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"170803772155648649000782583673804755121","date":"2025-07-09T18:38:59+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-09T12:05:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"128347686428509244395102090515412050452","date":"2025-07-09T11:31:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"284829189888652938208511187227264019668","date":"2025-07-09T08:39:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"66188692130192187129217063952532450907","date":"2025-07-09T01:34:55+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"9434957720985050454070631726175634515","date":"2025-07-08T19:47:10+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"85920557229480438966627487730296358554","date":"2025-07-08T19:39:02+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-08T19:37:16+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-08T19:18:22+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-05-21T02:54:47+00:00","index":"","fulltext":""},{"type":"submitted","content":"The Journal of Supercomputing","date":"2025-05-14T12:15:29+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"the-journal-of-supercomputing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [The Journal of Supercomputing](https://www.springer.com/journal/11227)","snPcode":"11227","submissionUrl":"https://submission.nature.com/new-submission/11227/3","title":"The Journal of Supercomputing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"046dd685-84fe-46e7-bd33-e5af3660e068","owner":[],"postedDate":"July 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2025-09-16T14:23:26+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-11 05:52:30","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6664114","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6664114","identity":"rs-6664114","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}