Dual-Branch Feature Continual Learning for Few-Shot Semantic Segmentation

Dual-Branch Feature Continual Learning for Few-Shot Semantic 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 Dual-Branch Feature Continual Learning for Few-Shot Semantic Segmentation Fuqiang Chen, Huifang Zhao, Guisheng Tan, Yijin Shi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8159882/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 30 Mar, 2026 Read the published version in Signal, Image and Video Processing → Version 1 posted 7 You are reading this latest preprint version Abstract To address the issue that large variations in appearance and scale among images of the same class lead to poor generalization of segmentation models on unseen images, a few-shot semantic segmentation method based on dual-branch feature continual learning is proposed. First, a pair of shared-weight backbone networks map the support and query images into a deep feature space. The ground truth mask of the support image is then used to separate its encoded features into foreground object features and background features. Next, the CLIP-encoded textual features are used to disentangle the mixed query features. The separated dual-branch foreground and background features are subsequently fused and aligned, enabling continual learning between branches through the target task. Finally, mask average pooling is applied to the fused foreground and background features to generate corresponding prototypes, and a parameter-free metric matching is performed to match query features with the prototype set on a per-pixel basis. Experiments on the PASCAL-5i and COCO-20i datasets under 1-shot and 5-shot settings demonstrate that the proposed method achieves competitive segmentation performance, comparable to state-of-the-art few-shot semantic segmentation methods. Semantic segmentation continual learning dual-branch features feature disentanglement mask average pooling Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 30 Mar, 2026 Read the published version in Signal, Image and Video Processing → Version 1 posted Editorial decision: Revision requested 28 Dec, 2025 Reviews received at journal 20 Dec, 2025 Reviewers agreed at journal 07 Dec, 2025 Reviewers invited by journal 04 Dec, 2025 Editor assigned by journal 20 Nov, 2025 Submission checks completed at journal 20 Nov, 2025 First submitted to journal 19 Nov, 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-8159882","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":556422264,"identity":"cd319ae0-eeb2-4adc-825c-9c3d79a845f7","order_by":0,"name":"Fuqiang Chen","email":"","orcid":"","institution":"Lanzhou University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Fuqiang","middleName":"","lastName":"Chen","suffix":""},{"id":556422265,"identity":"a7559df9-785f-4bba-9d53-a1a2558849d2","order_by":1,"name":"Huifang Zhao","email":"","orcid":"","institution":"Xi’an Jiaotong University","correspondingAuthor":false,"prefix":"","firstName":"Huifang","middleName":"","lastName":"Zhao","suffix":""},{"id":556422267,"identity":"6e629f01-640f-43b3-bf6c-ceea05bb0451","order_by":2,"name":"Guisheng Tan","email":"","orcid":"","institution":"Lijiang Culture and Tourism College","correspondingAuthor":false,"prefix":"","firstName":"Guisheng","middleName":"","lastName":"Tan","suffix":""},{"id":556422268,"identity":"2c5e3634-84d5-48ef-903a-32cbe03ba2af","order_by":3,"name":"Yijin Shi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA70lEQVRIiWNgGAWjYDACCcYGhgQeCQZ+BoYEhgqQyAFitUgCKYYzxGmB0gYglURpkZ/d3CbxQMYiz/hGwuMPB9sY5PhuJDB+LsCjxeDOwTYJoMOKzc4cSDAAajGWvJHALD0DnxaJRLCWxG3HGxKSP7YxJG64kcDGzIPPYTOgWjY3MyQcANpST1ALww2olg3sDYkNQC0JBoS0GNxIbLYAaZlx5kAyw4FzEoYzzzxslsbvsPSHN3/21CX2z8hJ/nCgzEae73jywc94HcbAwCLB2AOieRIYINEEjFwCgPkDww8QzX6AkMpRMApGwSgYoQAAhsdRyW2W/38AAAAASUVORK5CYII=","orcid":"","institution":"Xinjiang University of Political Science and Law","correspondingAuthor":true,"prefix":"","firstName":"Yijin","middleName":"","lastName":"Shi","suffix":""}],"badges":[],"createdAt":"2025-11-20 03:23:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8159882/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8159882/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s11760-026-05256-y","type":"published","date":"2026-03-30T15:58:45+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":97650684,"identity":"9c067f60-515e-4e56-acd4-eb5eb8027236","added_by":"auto","created_at":"2025-12-08 06:20:08","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":51650338,"visible":true,"origin":"","legend":"","description":"","filename":"DualBranchFeatureContinualLearningforFewShotSemanticSegmentation.docx","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/e7348dea924506d3c1a15ee0.docx"},{"id":97650667,"identity":"78c5f7ca-1c52-41a8-a124-a5a65fc276a1","added_by":"auto","created_at":"2025-12-08 06:20:06","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5638,"visible":true,"origin":"","legend":"","description":"","filename":"54f9ec46630e42a3b63a05d8fecc8f5c.json","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/77a9e264893994200f2099d4.json"},{"id":97674674,"identity":"4ab03ce8-add3-46e0-97fa-6d92bd57ebd2","added_by":"auto","created_at":"2025-12-08 09:43:48","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":156696,"visible":true,"origin":"","legend":"","description":"","filename":"54f9ec46630e42a3b63a05d8fecc8f5c1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/5bdf4ec3041f5b4e7650b895.xml"},{"id":97673965,"identity":"7b97e119-72a6-4502-8c87-781402c38173","added_by":"auto","created_at":"2025-12-08 09:42:00","extension":"emf","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":736628,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.emf","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/27953b4842689aa29ff388bf.emf"},{"id":97650672,"identity":"0999a99d-09ea-405a-8074-93d624235f11","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"emf","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":381984,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.emf","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/245fada73acd4217fb3635a7.emf"},{"id":97650669,"identity":"c57bd707-4c5f-4fc2-9cc7-904b31f5a5c2","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"emf","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":239860,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.emf","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/5d5d3cc40771b7c537aa843a.emf"},{"id":97674015,"identity":"1c093fa1-fadb-4e33-8038-af5cc7250e48","added_by":"auto","created_at":"2025-12-08 09:42:09","extension":"emf","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6319432,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.emf","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/ef996b4f31aeab0b4856a807.emf"},{"id":97674426,"identity":"261d752f-5eb4-4c1c-9f89-b236410b306c","added_by":"auto","created_at":"2025-12-08 09:43:15","extension":"emf","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6031968,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.emf","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/e7b2c39de52135b707ddeaac.emf"},{"id":97650674,"identity":"7a80f1e4-5946-4549-bf26-13d50970b56a","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"emf","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11878088,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.emf","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/68241575bb577c9e4cfdeaf3.emf"},{"id":97674098,"identity":"9abeab6d-a234-41bc-b7b3-06c2c7b83485","added_by":"auto","created_at":"2025-12-08 09:42:25","extension":"emf","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2794696,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.emf","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/4ea3ca2e4eb2e6be0c2730a6.emf"},{"id":97650670,"identity":"e61cb736-1d97-4a85-9735-fcf3e0eab050","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":32589,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/6af43459115a5f17a1f2bfb0.png"},{"id":97650671,"identity":"2e29d8ef-1997-4872-b385-95a62efe4b67","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11565,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/6095d252eb236d13e1793e1b.png"},{"id":97673052,"identity":"790cb887-62e6-46cc-ac5c-2d14ab3a67a4","added_by":"auto","created_at":"2025-12-08 09:39:20","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":10110,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/493bae5f138d5197b02dbd6f.png"},{"id":97673629,"identity":"fa42f77f-7aab-471f-b055-4a07391a575c","added_by":"auto","created_at":"2025-12-08 09:40:52","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":141733,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/d88ab807a4bdf3e27924b1f5.png"},{"id":97650683,"identity":"6f14270b-3a3e-4c88-a790-f8a940c9d191","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":169241,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/3e1950519ed0d64a0cde08e7.png"},{"id":97650681,"identity":"9bd6a0d2-8968-48d0-99dd-b9fe57e204ec","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":171076,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/27a01299f50a669e02b99a0b.png"},{"id":97650678,"identity":"fc991167-c51b-4304-993a-3871c73f71bb","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":106452,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/440b83ba2672ca0d134483e7.png"},{"id":97650675,"identity":"37407a0f-5202-41a8-aece-b0676fd8727f","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"xml","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":156220,"visible":true,"origin":"","legend":"","description":"","filename":"54f9ec46630e42a3b63a05d8fecc8f5c1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/f1fb9c1077d9125637d4e197.xml"},{"id":97650682,"identity":"0a407d9e-30ad-46a2-bb0a-0e972a85f783","added_by":"auto","created_at":"2025-12-08 06:20:07","extension":"html","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":167897,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1/25878c45d1bf6def488b89e1.html"},{"id":106343612,"identity":"7d1ae079-92d9-4af4-8f85-16d7fad89ab5","added_by":"auto","created_at":"2026-04-07 16:07:25","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":764640,"visible":true,"origin":"","legend":"","description":"","filename":"DualBranchFeatureContinualLearningforFewShotSemanticSegmentation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8159882/v1_covered_49b2173f-2930-4271-ac49-63bcc048646d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Dual-Branch Feature Continual Learning for Few-Shot Semantic 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":"signal-image-and-video-processing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"sivp","sideBox":"Learn more about [Signal, Image and Video Processing](http://link.springer.com/journal/11760)","snPcode":"11760","submissionUrl":"https://submission.nature.com/new-submission/11760/3","title":"Signal, Image and Video Processing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Semantic segmentation, continual learning, dual-branch features, feature disentanglement, mask average pooling","lastPublishedDoi":"10.21203/rs.3.rs-8159882/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8159882/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTo address the issue that large variations in appearance and scale among images of the same class lead to poor generalization of segmentation models on unseen images, a few-shot semantic segmentation method based on dual-branch feature continual learning is proposed. First, a pair of shared-weight backbone networks map the support and query images into a deep feature space. The ground truth mask of the support image is then used to separate its encoded features into foreground object features and background features. Next, the CLIP-encoded textual features are used to disentangle the mixed query features. The separated dual-branch foreground and background features are subsequently fused and aligned, enabling continual learning between branches through the target task. Finally, mask average pooling is applied to the fused foreground and background features to generate corresponding prototypes, and a parameter-free metric matching is performed to match query features with the prototype set on a per-pixel basis. Experiments on the PASCAL-5i and COCO-20i datasets under 1-shot and 5-shot settings demonstrate that the proposed method achieves competitive segmentation performance, comparable to state-of-the-art few-shot semantic segmentation methods.\u003c/p\u003e","manuscriptTitle":"Dual-Branch Feature Continual Learning for Few-Shot Semantic Segmentation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-08 06:20:02","doi":"10.21203/rs.3.rs-8159882/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-12-28T15:59:28+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-21T03:13:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"57314254804891406529459854085039737450","date":"2025-12-07T08:38:53+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-12-04T21:26:52+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-11-20T13:17:02+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-20T13:16:06+00:00","index":"","fulltext":""},{"type":"submitted","content":"Signal, Image and Video Processing","date":"2025-11-20T03:14:33+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"signal-image-and-video-processing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"sivp","sideBox":"Learn more about [Signal, Image and Video Processing](http://link.springer.com/journal/11760)","snPcode":"11760","submissionUrl":"https://submission.nature.com/new-submission/11760/3","title":"Signal, Image and Video Processing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"4de7a0fa-6b89-4e19-bcca-7b78882eb0f2","owner":[],"postedDate":"December 8th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-04-07T16:03:33+00:00","versionOfRecord":{"articleIdentity":"rs-8159882","link":"https://doi.org/10.1007/s11760-026-05256-y","journal":{"identity":"signal-image-and-video-processing","isVorOnly":false,"title":"Signal, Image and Video Processing"},"publishedOn":"2026-03-30 15:58:45","publishedOnDateReadable":"March 30th, 2026"},"versionCreatedAt":"2025-12-08 06:20:02","video":"","vorDoi":"10.1007/s11760-026-05256-y","vorDoiUrl":"https://doi.org/10.1007/s11760-026-05256-y","workflowStages":[]},"version":"v1","identity":"rs-8159882","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8159882","identity":"rs-8159882","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}