Towards Robust Federated Test-Time Adaptation: Dynamic Client Collaboration and Category-Aware Uncertainty

Towards Robust Federated Test-Time Adaptation: Dynamic Client Collaboration and Category-Aware Uncertainty | 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 Towards Robust Federated Test-Time Adaptation: Dynamic Client Collaboration and Category-Aware Uncertainty Yongcai Li, Yuexia Zhou, Xiangyu Liu, Kai Chen, Jinpeng Chen, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9453712/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Federated test-time adaptation (FTTA) enables privacy-preserving model adaptation to unlabeled target data during inference, yet it struggles with dynamic source client availability and uncertain test samples under distribution shifts. Existing methods overlook offline client impacts and rely on noisy pseudo-labels, leading to error accumulation. To address these challenges, we propose a novel FTTA framework, termed \underline{Fed}erated test-time adaptation under \underline{D}ynamic client collaboration and \underline{C}ategory-aware uncertainty (FedDC), which integrates dynamic client weighting initialization and class-aware margin thresholds to improve adaptation robustness. During source training, clients are aggregated adaptively based on participation history to reduce bias. At test time, category-specific thresholds separate confident and uncertain samples, preserving prediction uncertainty to mitigate noise. Extensive experiments on CIFAR100, Tiny-ImageNet, PACS, and CarlaTTA show that FedDC outperforms baseline methods under various distribution shifts, with clear improvements in adaptation accuracy and stability. The proposed design supports robust deployment in real-world decentralized visual systems. The source code is publicly available at https://github.com/ycarobot/FedDC. Software Engineering Image classification federated learning test-time adaptation class-wise pseudo-label distribution shift Full Text Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted 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-9453712","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":625278311,"identity":"0f55c1c0-b069-40f7-a34b-b8809a4ef2e2","order_by":0,"name":"Yongcai Li","email":"","orcid":"","institution":"Foshan University","correspondingAuthor":false,"prefix":"","firstName":"Yongcai","middleName":"","lastName":"Li","suffix":""},{"id":625278312,"identity":"00a9603b-328e-44f1-be3a-80da71b7e396","order_by":1,"name":"Yuexia Zhou","email":"","orcid":"","institution":"Foshan University","correspondingAuthor":false,"prefix":"","firstName":"Yuexia","middleName":"","lastName":"Zhou","suffix":""},{"id":625278313,"identity":"94612e3e-bd7f-4a95-9ec2-e2f4531c45c0","order_by":2,"name":"Xiangyu Liu","email":"","orcid":"","institution":"Foshan University","correspondingAuthor":false,"prefix":"","firstName":"Xiangyu","middleName":"","lastName":"Liu","suffix":""},{"id":625278314,"identity":"3a97b5f2-6f7a-434d-b390-0c73a402702f","order_by":3,"name":"Kai Chen","email":"","orcid":"","institution":"Guangzhou University","correspondingAuthor":false,"prefix":"","firstName":"Kai","middleName":"","lastName":"Chen","suffix":""},{"id":625278315,"identity":"d76d752e-e595-4408-bca2-ce0817978d0c","order_by":4,"name":"Jinpeng Chen","email":"","orcid":"","institution":"Foshan University","correspondingAuthor":false,"prefix":"","firstName":"Jinpeng","middleName":"","lastName":"Chen","suffix":""},{"id":625278316,"identity":"4ecd7d85-cdae-4ded-bcf7-25b69488f35d","order_by":5,"name":"Chang'an Yi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7klEQVRIiWNgGAWjYDACZjACsw58gDASiNbCljiDOC0McC08hsRpMTjOwPi5oOaOXf/sno8NH/4cZuBnzzFg+LkDj5bDDMzSM449S55x5+zGxplthxkke94YMPaewauFjZmH7XAyw43c7Y95Gw4zGNzIMWBmbCOk5d/hZPkbOQ+beYAOsydKC2/bYTug4YzNQOsYDCQIaJE8zNgszdt3OMHwRpoh0C/pPBJnnhUc7MWjhe/84YOfeb4dtpe7kfwQGGLWcvztyRsf/MSjReEAYwOITmyACvCAiAO4NTAwyEOV2uNTNApGwSgYBSMcAAAp+VOm++Q5ZAAAAABJRU5ErkJggg==","orcid":"","institution":"Foshan University","correspondingAuthor":true,"prefix":"","firstName":"Chang'an","middleName":"","lastName":"Yi","suffix":""}],"badges":[],"createdAt":"2026-04-18 03:12:46","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9453712/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9453712/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107387684,"identity":"35d0851f-2cf6-4104-b21b-c3be45a33319","added_by":"auto","created_at":"2026-04-21 04:11:08","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":28792292,"visible":true,"origin":"","legend":"","description":"","filename":"FedDCMainManuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9453712/v1_covered_fce11ee5-e6e9-4265-aa75-d6bd97e89ca7.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eTowards Robust Federated Test-Time Adaptation: Dynamic Client Collaboration and Category-Aware Uncertainty\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Foshan University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Image classification, federated learning, test-time adaptation, class-wise pseudo-label, distribution shift","lastPublishedDoi":"10.21203/rs.3.rs-9453712/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9453712/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eFederated test-time adaptation (FTTA) enables privacy-preserving model adaptation to unlabeled target data during inference, yet it struggles with dynamic source client availability and uncertain test samples under distribution shifts. Existing methods overlook offline client impacts and rely on noisy pseudo-labels, leading to error accumulation. To address these challenges, we propose a novel FTTA framework, termed \\underline{Fed}erated test-time adaptation under \\underline{D}ynamic client collaboration and \\underline{C}ategory-aware uncertainty (FedDC), which integrates dynamic client weighting initialization and class-aware margin thresholds to improve adaptation robustness. During source training, clients are aggregated adaptively based on participation history to reduce bias. At test time, category-specific thresholds separate confident and uncertain samples, preserving prediction uncertainty to mitigate noise. Extensive experiments on CIFAR100, Tiny-ImageNet, PACS, and CarlaTTA show that FedDC outperforms baseline methods under various distribution shifts, with clear improvements in adaptation accuracy and stability. The proposed design supports robust deployment in real-world decentralized visual systems. The source code is publicly available at https://github.com/ycarobot/FedDC.\u003c/p\u003e","manuscriptTitle":"Towards Robust Federated Test-Time Adaptation: Dynamic Client Collaboration and Category-Aware Uncertainty","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-21 04:10:33","doi":"10.21203/rs.3.rs-9453712/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"260b2b31-7bf5-424d-b22b-6b7e9f1d591f","owner":[],"postedDate":"April 21st, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":66554094,"name":"Software Engineering"}],"tags":[],"updatedAt":"2026-04-21T04:10:33+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-21 04:10:33","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9453712","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9453712","identity":"rs-9453712","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}