Multimodal Model Based on Contrastive Language-Image Pretraining for Micro-Expression Recognition

Multimodal Model Based on Contrastive Language-Image Pretraining for Micro-Expression Recognition | 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 Multimodal Model Based on Contrastive Language-Image Pretraining for Micro-Expression Recognition Peng Yang, Xiaoguang Wu, Yanyang Zhou, Qilin Wei, Zhifeng Zeng This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8133344/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 Recognizing involuntary, low-intensity micro-expressions (MEs) is challenging due to their subtlety and a lack of large-scale annotated data. This study introduces MECLIP, a novel dual-modal framework based on Contrastive Language-Image Pretraining (CLIP), to enhance ME recognition accuracy through enriched semantic supervision. MECLIP reformulates the CLIP architecture by incorporating a hierarchical temporal transformer to model visual dynamics and leverages a large language model to generate fine-grained physiological descriptors as textual guidance. An adaptive weighting mechanism fuses these spatiotemporal visual features with the nuanced textual semantics via contrastive learning. On the CAS(ME)³ dataset, MECLIP achieved a state-of-the-art unweighted average recall (UAR) of 39.6% and an unweighted F1-score (UF1) of 40.0%, outperforming existing benchmarks. The model also demonstrated strong zero-shot learning capabilities on the CASMEII dataset. Language-augmented multimodal learning presents a promising paradigm for improving micro-expression analysis, effectively compensating for data scarcity through fine-grained semantic feature alignment. Micro-expression recognition multimodal zero-shot learning CLIP 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-8133344","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":560522180,"identity":"5bd736e0-0ec0-4eed-8d8c-35c237388d3e","order_by":0,"name":"Peng Yang","email":"","orcid":"","institution":"Zhuhai College of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Peng","middleName":"","lastName":"Yang","suffix":""},{"id":560522181,"identity":"39325d16-5c42-49fe-ace4-1faa0b583cbe","order_by":1,"name":"Xiaoguang Wu","email":"","orcid":"","institution":"Zhuhai College of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Xiaoguang","middleName":"","lastName":"Wu","suffix":""},{"id":560522182,"identity":"72018209-45cd-4d30-985b-59267fa2ca93","order_by":2,"name":"Yanyang Zhou","email":"","orcid":"","institution":"Zhuhai College of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Yanyang","middleName":"","lastName":"Zhou","suffix":""},{"id":560522192,"identity":"f2137127-1912-4652-a5bf-83bb1f37b854","order_by":3,"name":"Qilin Wei","email":"","orcid":"","institution":"Zhuhai College of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Qilin","middleName":"","lastName":"Wei","suffix":""},{"id":560522197,"identity":"b86e2f56-4961-4996-9327-2c50c3a5fd98","order_by":4,"name":"Zhifeng Zeng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAr0lEQVRIiWNgGAWjYFACxocPoCwDYrUwG4OUSpCkxUyCNC3y/YfZKn/U1NUxsDdvk2CouUNYC2PDYbYbEscOSzDwHCuTYDj2jAhnMfYfu2HYcECCQSLHTAJoAGEtbMzMbAWJDXUSDPJviNTCw8bMxnCwgRloCw+RWiR4mJklG44dlmzjSSu2SDhGhBZgiDF+BIYYPz/74Y03PtQQoQUO2EBEAgkaRsEoGAWjYBTgAQCRIy4QDcNd9QAAAABJRU5ErkJggg==","orcid":"","institution":"Zhuhai College of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Zhifeng","middleName":"","lastName":"Zeng","suffix":""}],"badges":[],"createdAt":"2025-11-17 09:23:34","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8133344/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8133344/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":98441520,"identity":"098d6639-cbe3-43b8-95f5-d5da935f3725","added_by":"auto","created_at":"2025-12-17 17:05:34","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6467607,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/b00af6a8a0a014644b946706.docx"},{"id":98396867,"identity":"33f0f20c-f0d5-47f6-981b-13e1d672c307","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6560,"visible":true,"origin":"","legend":"","description":"","filename":"dac28a081eee4a0abadc0b851046a1d7.json","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/b547965821f74ae872615a0f.json"},{"id":98441742,"identity":"ba4f7771-a2a2-4b97-a175-8ca8770efc24","added_by":"auto","created_at":"2025-12-17 17:05:45","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":138314,"visible":true,"origin":"","legend":"","description":"","filename":"dac28a081eee4a0abadc0b851046a1d71enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/4a52b9c7b8431f854fda0796.xml"},{"id":98396865,"identity":"fe0a45a7-8c44-4415-b5c3-7acb3ed8f4b2","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":121186,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/5b8d0f7dc0c096a14ae7fab0.png"},{"id":98396863,"identity":"471fbf1f-1463-4195-ad6c-059a2c9ce72f","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":160946,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/818d35b0cb3cd6ca5a93dea5.png"},{"id":98440862,"identity":"a45f0b9d-fd80-4691-b063-1e62a7326633","added_by":"auto","created_at":"2025-12-17 17:04:31","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":63841,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/4c70fe996a000475ac5d13f5.png"},{"id":98396872,"identity":"5dcd93b5-274d-4102-8e99-28bfafed84e8","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":70565,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/f06c0881cb5d0222d0bb7218.png"},{"id":98440472,"identity":"f5c13489-eb55-456f-a80e-8bd7c90724b4","added_by":"auto","created_at":"2025-12-17 17:03:54","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":26885,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/23e5c7c00c52655363d620d1.png"},{"id":98396868,"identity":"57217ded-913f-4b6a-a110-ae9edbb69a27","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1106767,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/31bea6c910e00f5910dbfa9d.png"},{"id":98396873,"identity":"d8c6de96-6c0f-4845-a0b5-888090461dfe","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1112918,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/1ec790adc7c5c62706a1ea9a.png"},{"id":98396880,"identity":"3f60f6b8-6b0a-4ac4-8112-ed5f5da69029","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3731185,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/677fdf6154c8385568c8e64c.png"},{"id":98440824,"identity":"62f2be98-d679-4081-a34b-fac196084e0f","added_by":"auto","created_at":"2025-12-17 17:04:25","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25454,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/f7983da535dc1b23d826183f.png"},{"id":98440493,"identity":"06864b95-a0de-4879-942f-c30e96506c0a","added_by":"auto","created_at":"2025-12-17 17:03:56","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":33048,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/570df84394986214b6861754.png"},{"id":98439747,"identity":"5e75aa1a-9396-4fa0-ae6e-3ca9b1299eb7","added_by":"auto","created_at":"2025-12-17 17:02:52","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":14170,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/22e43c88d6b7d38280603cf6.png"},{"id":98440458,"identity":"868ef787-0363-4ec9-8753-18b5ce9a1532","added_by":"auto","created_at":"2025-12-17 17:03:53","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":21643,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/c07bed7ff0f597882696a0af.png"},{"id":98439635,"identity":"4db4889f-387a-4eea-b39d-a599795c3965","added_by":"auto","created_at":"2025-12-17 17:02:16","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8612,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/4b4a1eddc5d7b97d55053c58.png"},{"id":98396877,"identity":"9d8d2fed-40c5-4678-97b6-eb2565949327","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":148473,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/4e17c49e27eac02d50922dbf.png"},{"id":98396881,"identity":"7eb01feb-963c-4b6b-aa42-4f2f57c61231","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":125719,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/fa228074fec3d26521f80f46.png"},{"id":98440979,"identity":"d219a7b3-f978-454e-a073-ce8a4cd56141","added_by":"auto","created_at":"2025-12-17 17:04:43","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":841074,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/fa4a947568b078ed8995c08f.png"},{"id":98440592,"identity":"8bf37a8f-0541-4cef-8364-3648c22a45ac","added_by":"auto","created_at":"2025-12-17 17:04:04","extension":"xml","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":135104,"visible":true,"origin":"","legend":"","description":"","filename":"dac28a081eee4a0abadc0b851046a1d71structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/a095bc7e9b25d0ad741e06b9.xml"},{"id":98396883,"identity":"5748cdd5-e592-42c5-b2b7-30df31a5a642","added_by":"auto","created_at":"2025-12-17 10:46:36","extension":"html","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":154705,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1/c923e1ce6e6c2491cf3f4206.html"},{"id":104812850,"identity":"1b9d298d-0727-48d4-97db-8eb785e892a3","added_by":"auto","created_at":"2026-03-17 13:05:09","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":880015,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8133344/v1_covered_69174476-5215-4f5c-9f64-590f875903bf.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Multimodal Model Based on Contrastive Language-Image Pretraining for Micro-Expression Recognition","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":"Micro-expression recognition, multimodal, zero-shot learning, CLIP","lastPublishedDoi":"10.21203/rs.3.rs-8133344/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8133344/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eRecognizing involuntary, low-intensity micro-expressions (MEs) is challenging due to their subtlety and a lack of large-scale annotated data. This study introduces MECLIP, a novel dual-modal framework based on Contrastive Language-Image Pretraining (CLIP), to enhance ME recognition accuracy through enriched semantic supervision. MECLIP reformulates the CLIP architecture by incorporating a hierarchical temporal transformer to model visual dynamics and leverages a large language model to generate fine-grained physiological descriptors as textual guidance. An adaptive weighting mechanism fuses these spatiotemporal visual features with the nuanced textual semantics via contrastive learning. On the CAS(ME)\u0026sup3; dataset, MECLIP achieved a state-of-the-art unweighted average recall (UAR) of 39.6% and an unweighted F1-score (UF1) of 40.0%, outperforming existing benchmarks. The model also demonstrated strong zero-shot learning capabilities on the CASMEII dataset. Language-augmented multimodal learning presents a promising paradigm for improving micro-expression analysis, effectively compensating for data scarcity through fine-grained semantic feature alignment.\u003c/p\u003e","manuscriptTitle":"Multimodal Model Based on Contrastive Language-Image Pretraining for Micro-Expression Recognition","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-17 10:46:27","doi":"10.21203/rs.3.rs-8133344/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":"84f81f66-ea76-4553-a6b1-6a7dd9f9f4cb","owner":[],"postedDate":"December 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-17T12:56:36+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-17 10:46:27","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8133344","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8133344","identity":"rs-8133344","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}