Multi-Teacher Knowledge Distillation via Tucker-Guided Representation Alignment and Adaptive Feature Mapping

Multi-Teacher Knowledge Distillation via Tucker-Guided Representation Alignment and Adaptive Feature Mapping | 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 Multi-Teacher Knowledge Distillation via Tucker-Guided Representation Alignment and Adaptive Feature Mapping Majid Sepahvand, Maytham N. Meqdad, Fardin Abdali Mohammadi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7184025/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 The knowledge distillation fused with the feature maps of multiple advanced teachers is a promising technique for training a student. However, if there is a high variance between the spatial shapes of the feature maps from teachers and students, effective distillation becomes challenging and it is wrong to directly calculate their differences. In this paper, a novel framework for structured knowledge distillation based on adaptive feature alignment and Tucker decomposition is proposed. The proposed framework uses a new combination of Tucker decomposition and learnable convolutional regression to enable structured, multi-path feature distillation from multiple teachers. The high-level feature tensor of each teacher is decomposed into core semantic representations, which are adaptively projected to student layers through learnable regressors. By providing semantically rich representations to guide several layers of the student network, the approach facilitates multi-teacher supervision. Finally, an adaptive hybrid loss is proposed to guide the transfer of core tensor knowledge from the teacher to the student. Experimental results on CIFAR-100 and Tiny-ImageNet demonstrate that our approach consistently outperforms state-of-the-art distillation baselines. According to the experimental results, the proposed knowledge distillation model achieved an accuracy of 96.48% and 91.70% for the classification images in the CIFAR-100 and Tiny-ImageNet datasets, respectively. Knowledge Distillation Tucker Decomposition Multi-Teacher Learning Feature Alignment Deep Neural Networks Full Text Additional Declarations No competing interests reported. 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-7184025","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":490073282,"identity":"2737d911-7189-4224-aeaa-3e47ccdf04f4","order_by":0,"name":"Majid Sepahvand","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6klEQVRIiWNgGAWjYPCCAyDC8AGQMIAIGBCnxdiAZC1mEkSp1W1gf/i4oOZOYv/s5m0VP/fYGRscYH74gaHgHk4tZgd4jI1nHHuWOOPOsbKbPc+SzQwOsBlLMBgU49Zy/w2bNA/bYWOGGzlmN3gOMNsYHGAwA7ovAY8t7M9/8/w7bCwP1FL450A9UAv7NwJaGMyYedsOyxkAtTDzHDgMdBgPIVt4jKVn9h2WM7yRViwtc+C4seRhnmKJBPwOe/i54NthHrkbyRs/vjlQbdh3vH3jhw9/cGsBAWZMLn4NGFpGwSgYBaNgFKABAMhuVDaoHoWBAAAAAElFTkSuQmCC","orcid":"","institution":"Arak University","correspondingAuthor":true,"prefix":"","firstName":"Majid","middleName":"","lastName":"Sepahvand","suffix":""},{"id":490073283,"identity":"757ce672-f42f-44e3-9020-2d0a7b277d3a","order_by":1,"name":"Maytham N. Meqdad","email":"","orcid":"","institution":"Al-Mustaqbal University College","correspondingAuthor":false,"prefix":"","firstName":"Maytham","middleName":"N.","lastName":"Meqdad","suffix":""},{"id":490073284,"identity":"c5623532-c3c6-490a-bdc6-bcda9ab5d37d","order_by":2,"name":"Fardin Abdali Mohammadi","email":"","orcid":"","institution":"Razi University","correspondingAuthor":false,"prefix":"","firstName":"Fardin","middleName":"Abdali","lastName":"Mohammadi","suffix":""}],"badges":[],"createdAt":"2025-07-22 07:43:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7184025/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7184025/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":89154281,"identity":"cf079e68-f9ac-4d99-9e05-d20e985badaa","added_by":"auto","created_at":"2025-08-15 13:47:04","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4624284,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7184025/v1_covered_1d3d78a7-b5b7-4fa9-99c9-376055722598.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Multi-Teacher Knowledge Distillation via Tucker-Guided Representation Alignment and Adaptive Feature Mapping","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","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":"Knowledge Distillation, Tucker Decomposition, Multi-Teacher Learning, Feature Alignment, Deep Neural Networks","lastPublishedDoi":"10.21203/rs.3.rs-7184025/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7184025/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The knowledge distillation fused with the feature maps of multiple advanced teachers is a promising technique for training a student. However, if there is a high variance between the spatial shapes of the feature maps from teachers and students, effective distillation becomes challenging and it is wrong to directly calculate their differences. In this paper, a novel framework for structured knowledge distillation based on adaptive feature alignment and Tucker decomposition is proposed. The proposed framework uses a new combination of Tucker decomposition and learnable convolutional regression to enable structured, multi-path feature distillation from multiple teachers. The high-level feature tensor of each teacher is decomposed into core semantic representations, which are adaptively projected to student layers through learnable regressors. By providing semantically rich representations to guide several layers of the student network, the approach facilitates multi-teacher supervision. Finally, an adaptive hybrid loss is proposed to guide the transfer of core tensor knowledge from the teacher to the student. Experimental results on CIFAR-100 and Tiny-ImageNet demonstrate that our approach consistently outperforms state-of-the-art distillation baselines. According to the experimental results, the proposed knowledge distillation model achieved an accuracy of 96.48% and 91.70% for the classification images in the CIFAR-100 and Tiny-ImageNet datasets, respectively.","manuscriptTitle":"Multi-Teacher Knowledge Distillation via Tucker-Guided Representation Alignment and Adaptive Feature Mapping","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-25 03:39:47","doi":"10.21203/rs.3.rs-7184025/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":"47f469e3-9267-4163-8ae3-345ef46908ad","owner":[],"postedDate":"July 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-08-15T13:38:56+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-25 03:39:47","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7184025","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7184025","identity":"rs-7184025","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}