Research on Intelligent Periodization and Imbalanced Classification of  Yungang Grottoes Buddha Heads Based on Deep Contrastive Learning

Research on Intelligent Periodization and Imbalanced Classification of Yungang Grottoes Buddha Heads Based on Deep Contrastive Learning | 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 Article Research on Intelligent Periodization and Imbalanced Classification of Yungang Grottoes Buddha Heads Based on Deep Contrastive Learning xitao wang, rui hu, yuewen hu, qingyang Li, Hanyue Zhang, Deng Pan, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9123382/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 historical periodization and classification of Buddha heads in the Yungang Grottoes is a crucial subject in cultural heritage conservation and archaeological research. Accurate periodization recognition is of great significance for grotto protection planning, disease monitoring, and restoration decision-making. Traditional periodization methods mainly rely on expert experience, which are highly subjective and inefficient, making it difficult to meet the practical needs of large-scale cultural relic protection. To address the class imbalance problem in the periodization and classification of Yungang Grottoes Buddha heads, this study proposes a deep learning method called ADNIC (Adaptive Dual-Branch Neural Framework for Imbalance Classification), which combines supervised contrastive learning with adaptive Focal Loss. This method adopts a dual-branch network architecture and a two-stage training strategy, effectively improving the recognition ability for minority classes. Experimental results show that the overall accuracy of this method reaches 90.3%, and the F1 score of the third period (minority class) increases from 0.364 to 0.833, significantly ameliorating the inadequate recognition of rare categories by traditional methods. This research provides technical support for the digital protection and intelligent management of the Yungang Grottoes, and also offers a referenceable method for the intelligent recognition and protection of other grotto temple cultural relics. Yungang Grottoes Buddha head images Supervised contrastive learning Class imbalance Deep learning Two-stage training 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-9123382","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":617580083,"identity":"7181bcbc-b8ee-4fa5-8f40-76841f72cee7","order_by":0,"name":"xitao wang","email":"","orcid":"","institution":"Shanghai University","correspondingAuthor":false,"prefix":"","firstName":"xitao","middleName":"","lastName":"wang","suffix":""},{"id":617580084,"identity":"776b3b16-ca3d-40f0-bcc7-126b0e7d40cd","order_by":1,"name":"rui hu","email":"","orcid":"","institution":"Shanghai University","correspondingAuthor":false,"prefix":"","firstName":"rui","middleName":"","lastName":"hu","suffix":""},{"id":617580085,"identity":"6d22c52f-0dc3-4053-b8b5-0d35bfbb6f47","order_by":2,"name":"yuewen hu","email":"","orcid":"","institution":"Shanghai University","correspondingAuthor":false,"prefix":"","firstName":"yuewen","middleName":"","lastName":"hu","suffix":""},{"id":617580087,"identity":"fcab6ca1-f0a5-4d44-bb31-97e3a0de8866","order_by":3,"name":"qingyang Li","email":"","orcid":"","institution":"Shanghai University","correspondingAuthor":false,"prefix":"","firstName":"qingyang","middleName":"","lastName":"Li","suffix":""},{"id":617580090,"identity":"6e5d6788-5fdd-4427-9dad-80875792df78","order_by":4,"name":"Hanyue Zhang","email":"","orcid":"","institution":"Shanghai University","correspondingAuthor":false,"prefix":"","firstName":"Hanyue","middleName":"","lastName":"Zhang","suffix":""},{"id":617580091,"identity":"0bb70004-7d86-429f-8aa0-74fd025b9f16","order_by":5,"name":"Deng Pan","email":"","orcid":"","institution":"Shanghai University","correspondingAuthor":false,"prefix":"","firstName":"Deng","middleName":"","lastName":"Pan","suffix":""},{"id":617580092,"identity":"16609a13-ab8c-4a68-a7e5-ac10a75874fb","order_by":6,"name":"Yue Zhang","email":"","orcid":"","institution":"Shanghai University","correspondingAuthor":false,"prefix":"","firstName":"Yue","middleName":"","lastName":"Zhang","suffix":""},{"id":617580093,"identity":"d74b0ee6-59af-4bc1-ba87-61cabba8afcb","order_by":7,"name":"Huakang Bian","email":"","orcid":"","institution":"Shanghai University","correspondingAuthor":false,"prefix":"","firstName":"Huakang","middleName":"","lastName":"Bian","suffix":""},{"id":617580095,"identity":"08aa539c-3d4a-4e00-9471-91375ec099db","order_by":8,"name":"Shunbo Hu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAyUlEQVRIiWNgGAWjYDACCQYGZgaGAwz8DDwgLjMJWiQbSNZicIBYLfKze8weF1TcSdx8/uwxCYYK68QG9rMH8GoxuHPG3HjGmWeJ227kpUkwnElPbODJS8CvRSLHTJq37TBQC4+ZBCOQ0SDBY4DfYTOgWjb3nwFq+UeEFoYbUC0bGHKAWhqI0GJwI61MesaZw8YzbuQYWyQcSzdu48kh5LDkbdIFFYdl+/vPGN74UGMt289+hoDDUEACELORoH4UjIJRMApGAQ4AADAoREUuPL2lAAAAAElFTkSuQmCC","orcid":"","institution":"Shanghai University","correspondingAuthor":true,"prefix":"","firstName":"Shunbo","middleName":"","lastName":"Hu","suffix":""},{"id":617580097,"identity":"16ce82b7-c98f-4d16-903e-d941670a0794","order_by":9,"name":"Jizhong Huang","email":"","orcid":"","institution":"Shanghai University","correspondingAuthor":false,"prefix":"","firstName":"Jizhong","middleName":"","lastName":"Huang","suffix":""}],"badges":[],"createdAt":"2026-03-14 14:40:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9123382/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9123382/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107707690,"identity":"b96c5bf6-47b0-4168-9b5f-d8acff86047e","added_by":"auto","created_at":"2026-04-24 09:20:55","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1471272,"visible":true,"origin":"","legend":"","description":"","filename":"article0321.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9123382/v1_covered_0d35f817-0b05-46d6-836e-2f9a86c82c60.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Research on Intelligent Periodization and Imbalanced Classification of Yungang Grottoes Buddha Heads Based on Deep Contrastive Learning","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":"Yungang Grottoes, Buddha head images, Supervised contrastive learning, Class imbalance, Deep learning, Two-stage training","lastPublishedDoi":"10.21203/rs.3.rs-9123382/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9123382/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The historical periodization and classification of Buddha heads in the Yungang Grottoes is a crucial subject in cultural heritage conservation and archaeological research. Accurate periodization recognition is of great significance for grotto protection planning, disease monitoring, and restoration decision-making. Traditional periodization methods mainly rely on expert experience, which are highly subjective and inefficient, making it difficult to meet the practical needs of large-scale cultural relic protection. To address the class imbalance problem in the periodization and classification of Yungang Grottoes Buddha heads, this study proposes a deep learning method called ADNIC (Adaptive Dual-Branch Neural Framework for Imbalance Classification), which combines supervised contrastive learning with adaptive Focal Loss. This method adopts a dual-branch network architecture and a two-stage training strategy, effectively improving the recognition ability for minority classes. Experimental results show that the overall accuracy of this method reaches 90.3\\%, and the F1 score of the third period (minority class) increases from 0.364 to 0.833, significantly ameliorating the inadequate recognition of rare categories by traditional methods. This research provides technical support for the digital protection and intelligent management of the Yungang Grottoes, and also offers a referenceable method for the intelligent recognition and protection of other grotto temple cultural relics.","manuscriptTitle":"Research on Intelligent Periodization and Imbalanced Classification of Yungang Grottoes Buddha Heads Based on Deep Contrastive Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-06 05:55:04","doi":"10.21203/rs.3.rs-9123382/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":"cf7def83-5169-4b86-97f1-16ac491767b3","owner":[],"postedDate":"April 6th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-23T15:11:16+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-06 05:55:04","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9123382","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9123382","identity":"rs-9123382","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}