FCL: Frequency-based Contrastive Learning for Generalizable Face Forgery Detection

FCL: Frequency-based Contrastive Learning for Generalizable Face Forgery Detection | 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 FCL: Frequency-based Contrastive Learning for Generalizable Face Forgery Detection Yu Zhu, Shengze Wang, Yufeng Gu, Ziming Zhu, Nan Wang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8126610/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract With the proliferation of deepfake techniques, face forgery detection has emerged as a critical area of research to mitigate the risks associated with facial manipulation. Recent studies have highlighted the superiority of frequency information over color-space input, especially in high-compression scenarios. However, most frequency-enhanced methods employ a two-stream network architecture, assigning separate encoders for RGB and frequency domains, and rely solely on vanilla binary cross-entropy loss, limiting their generalization ability.To address these limitations, we propose a novel framework, Frequency-enhanced Contrastive Learning (FCL), which trains the model in a supervised manner using contrastive loss. FCL treats two views of different modalities generated from the same image as positive pairs and samples with opposite labels as negative pairs. This approach pulls features from the RGB and high-frequency domains closer while pushing features of pristine and forgery faces apart. Additionally, we introduce a Shallow Feature Supplement (SFS) module to complement local information from low-level shallow feature embeddings into high-level feature maps, and a Dual Modal Fusion (DMF) module to adaptively aggregate information from both domains.Extensive experiments on seven datasets demonstrate the superior generalization of our method compared to state-of-the-art competitors. Notably, FCL achieves a remarkable accuracy of 92.29% and an AUC of 94.25% on the heavily compressed FaceForensics++ dataset, showcasing its robustness against compression. These results underscore the potential of FCL as a powerful tool for face forgery detection in real-world scenarios.The source code ofour proposed algorithm are available at https://github.com/Y30230924Wang/Frequency-Enhanced-Contrastive-Learning . Face forgery detection frequency contrastive learning generalization ability Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 06 Feb, 2026 Reviews received at journal 02 Feb, 2026 Reviewers agreed at journal 30 Dec, 2025 Reviews received at journal 23 Dec, 2025 Reviewers agreed at journal 23 Dec, 2025 Reviewers invited by journal 23 Dec, 2025 Editor assigned by journal 16 Dec, 2025 Submission checks completed at journal 18 Nov, 2025 First submitted to journal 16 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-8126610","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":565008593,"identity":"a56c6f97-0fb1-425d-ae69-91eece12416a","order_by":0,"name":"Yu Zhu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzklEQVRIiWNgGAWjYDACCRBRAeOxEa3lDIhgJkULYxspWuRn95hJ886zkTc4f/4Aw4eywwz8sxvwa2GccwaoZVua4YYbyQyMM84dZpC4cwC/FmaJHJCWwwkGN5gZmHnbDjMYSCTg18IG1jLnf4LB+cMMzH+J0cID1tJwIMHgQDIDMyMxWiQk0oot5xxLNpx5I9ngYM+5dB6JGwS0yM9I3njjTY2dPN/5gw8f/CizluOfQUALELBIwFgHQC4lqB4ImD8Qo2oUjIJRMApGMAAAnVQ8R3MyPh0AAAAASUVORK5CYII=","orcid":"","institution":"East China University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Yu","middleName":"","lastName":"Zhu","suffix":""},{"id":565008594,"identity":"7eac37b8-9c03-4193-a7d3-7caed8fca4ef","order_by":1,"name":"Shengze Wang","email":"","orcid":"","institution":"East China University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Shengze","middleName":"","lastName":"Wang","suffix":""},{"id":565008595,"identity":"b8070c5f-e3df-4dd2-a4c6-cca2ff34f4c8","order_by":2,"name":"Yufeng Gu","email":"","orcid":"","institution":"East China University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Yufeng","middleName":"","lastName":"Gu","suffix":""},{"id":565008596,"identity":"6af2554c-ebfc-40a2-b3f3-eded2a60bf3b","order_by":3,"name":"Ziming Zhu","email":"","orcid":"","institution":"East China University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Ziming","middleName":"","lastName":"Zhu","suffix":""},{"id":565008597,"identity":"8a651ee7-0073-4acb-80e2-5faf9f789dc0","order_by":4,"name":"Nan Wang","email":"","orcid":"","institution":"East China University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Nan","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2025-11-16 10:23:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8126610/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8126610/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99311763,"identity":"2851c75f-1c19-44cf-98ac-74c5d0869fd9","added_by":"auto","created_at":"2025-12-31 16:16:53","extension":"json","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7053,"visible":true,"origin":"","legend":"","description":"","filename":"ad06baad691845be868fa8fcb893b92a.json","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/38863d2c9471a2c405779c0c.json"},{"id":99312057,"identity":"609c4ffe-a8d1-432f-a52b-464db07bd90f","added_by":"auto","created_at":"2025-12-31 16:17:49","extension":"xml","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":163843,"visible":true,"origin":"","legend":"","description":"","filename":"ad06baad691845be868fa8fcb893b92a1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/6cbdfd9b4c1cf462d78d816b.xml"},{"id":99312517,"identity":"92c6c8da-6834-493f-8f46-3db4d4a82671","added_by":"auto","created_at":"2025-12-31 16:19:03","extension":"aux","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17022,"visible":true,"origin":"","legend":"","description":"","filename":"FCL.aux","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/27702a6bb4478c8793cfc8bb.aux"},{"id":98982672,"identity":"16355ee6-2b61-47fd-af96-d765c40e6930","added_by":"auto","created_at":"2025-12-25 07:27:53","extension":"bbl","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29024,"visible":true,"origin":"","legend":"","description":"","filename":"FCL.bbl","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/19bb851661d37a18adc1f16a.bbl"},{"id":98982676,"identity":"803aed44-33db-4d88-ad3f-a314895c550d","added_by":"auto","created_at":"2025-12-25 07:27:53","extension":"blg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2770,"visible":true,"origin":"","legend":"","description":"","filename":"FCL.blg","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/989a963c7aafbe7aa12d1299.blg"},{"id":99312499,"identity":"faaeaa6f-279d-4ca5-9dce-4aca942f0495","added_by":"auto","created_at":"2025-12-31 16:19:00","extension":"fdb_latexmk","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":19996,"visible":true,"origin":"","legend":"","description":"","filename":"FCL.fdblatexmk","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/526d41681a04fe1f802091b7.fdb_latexmk"},{"id":98982670,"identity":"419f916f-7cb1-4f8e-aa25-32f1977f7352","added_by":"auto","created_at":"2025-12-25 07:27:53","extension":"fls","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64206,"visible":true,"origin":"","legend":"","description":"","filename":"FCL.fls","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/0c5128946a1c1dfe4705cf1c.fls"},{"id":98982675,"identity":"482d15b1-72a8-460b-bdd9-9492a8776074","added_by":"auto","created_at":"2025-12-25 07:27:53","extension":"log","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35016,"visible":true,"origin":"","legend":"","description":"","filename":"FCL.log","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/fb785d81cd456acfde991e3f.log"},{"id":98982689,"identity":"cf44f2d1-3941-4d7c-acd6-d17886a9363b","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"out","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3644,"visible":true,"origin":"","legend":"","description":"","filename":"FCL.out","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/71a8c2b38a0c91b5bb6ca162.out"},{"id":99311960,"identity":"6cf8bcae-2820-4b56-a4cd-24635474fe7c","added_by":"auto","created_at":"2025-12-31 16:17:30","extension":"pdf","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2450180,"visible":true,"origin":"","legend":"","description":"","filename":"FCL.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/ef26daa2549dad843c01870b.pdf"},{"id":98982688,"identity":"e70c755b-8ab6-480c-a586-a11173efd6bf","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"gz","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":208385,"visible":true,"origin":"","legend":"","description":"","filename":"FCL.synctex.gz","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/1aca04d91aff33cf5ceb4872.gz"},{"id":99312133,"identity":"0b25b120-7d11-47de-aef2-421822a6cf4d","added_by":"auto","created_at":"2025-12-31 16:18:09","extension":"pdf","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2456770,"visible":true,"origin":"","legend":"","description":"","filename":"FCLfixed.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/eaf48acf9e39bbe6feb4b633.pdf"},{"id":98982692,"identity":"3c1a10c9-e728-4483-a5cb-fc64ca47764c","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"pptx","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":531381,"visible":true,"origin":"","legend":"","description":"","filename":"figure1fixed.pptx","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/8c46dce4b3ac215037414b1b.pptx"},{"id":98982679,"identity":"709c6b08-543a-42a5-9791-b2ea87abfd98","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"pptx","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":918450,"visible":true,"origin":"","legend":"","description":"","filename":"figure2fixed.pptx","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/38119abc7ffdbb8026d83bb6.pptx"},{"id":98982687,"identity":"bd4224b6-5058-4886-aacf-03687a097b1b","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"pptx","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":591869,"visible":true,"origin":"","legend":"","description":"","filename":"figure3fixed.pptx","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/60124997d7650811e9bdccca.pptx"},{"id":98982683,"identity":"1409c141-5c4d-4db3-9478-4c4a58e87adc","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"pptx","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":74477,"visible":true,"origin":"","legend":"","description":"","filename":"figure4fixed.pptx","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/ed3c1484aabdc095ac5f096b.pptx"},{"id":99312394,"identity":"13779a8e-9ffa-47c0-967a-8d05a1811a89","added_by":"auto","created_at":"2025-12-31 16:18:56","extension":"pptx","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":574962,"visible":true,"origin":"","legend":"","description":"","filename":"figure5fixed.pptx","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/5e6768ea8aadfed590eddee8.pptx"},{"id":98982701,"identity":"fe7cdf37-3cf5-49da-8873-e99a0d591cd6","added_by":"auto","created_at":"2025-12-25 07:27:55","extension":"pptx","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5852439,"visible":true,"origin":"","legend":"","description":"","filename":"figure6fixed.pptx","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/67b5beb1d89d621293617eee.pptx"},{"id":98982681,"identity":"7d5cfbab-89cf-4760-b642-3e2a2f833fb0","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"pptx","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":977342,"visible":true,"origin":"","legend":"","description":"","filename":"figure7fixed.pptx","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/7379aa22584d643f42a1bb69.pptx"},{"id":98982699,"identity":"9f7f3a8d-7441-4f0a-a1c7-d1282c6ccf62","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"pdf","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":373835,"visible":true,"origin":"","legend":"","description":"","filename":"graphicabs.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/3a78367f8c8e9154c349f413.pdf"},{"id":98982684,"identity":"a5083a48-1d95-4c7f-8b51-926fc89e3207","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"bst","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":151345,"visible":true,"origin":"","legend":"","description":"","filename":"snapacite.bst","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/0e881544e1653b1365430827.bst"},{"id":98982700,"identity":"7712e56f-49cf-4f94-afe0-340b60369d97","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"bst","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29828,"visible":true,"origin":"","legend":"","description":"","filename":"snaps.bst","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/fe2534104bb002de820589e0.bst"},{"id":98982685,"identity":"835a213e-6704-4eab-89e2-1d1c73f7537e","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"bst","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35515,"visible":true,"origin":"","legend":"","description":"","filename":"snbasic.bst","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/38c8596e90e5542c1ba89c86.bst"},{"id":98982686,"identity":"71beb549-3758-4403-a765-ae5e76e772fc","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"bst","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":33968,"visible":true,"origin":"","legend":"","description":"","filename":"snchicago.bst","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/facce93a409d604175a35fe7.bst"},{"id":98982697,"identity":"45f27f05-b125-4e29-89e0-f47bbfb5464a","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"cls","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":57660,"visible":true,"origin":"","legend":"","description":"","filename":"snjnl.cls","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/3257db3083db810483438ecf.cls"},{"id":99311864,"identity":"dbaf08cc-a255-497f-b0b3-89588ff98867","added_by":"auto","created_at":"2025-12-31 16:17:10","extension":"bst","order_by":32,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64023,"visible":true,"origin":"","legend":"","description":"","filename":"snmathphysay.bst","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/3548d5865efba70572bf993b.bst"},{"id":99311758,"identity":"5ea38db4-4f73-4230-88ed-6f29ce7780e8","added_by":"auto","created_at":"2025-12-31 16:16:50","extension":"bst","order_by":33,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64164,"visible":true,"origin":"","legend":"","description":"","filename":"snmathphysnum.bst","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/227b3e4992b3103d968bc871.bst"},{"id":98982696,"identity":"7c3f360a-7d29-49b0-9c68-692632c03dd9","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"bst","order_by":34,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":39056,"visible":true,"origin":"","legend":"","description":"","filename":"snnature.bst","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/2abf79975f44fbfde0defd91.bst"},{"id":98982694,"identity":"78b785eb-e04b-47ba-9d06-e9a6789ec54e","added_by":"auto","created_at":"2025-12-25 07:27:54","extension":"bst","order_by":35,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":39951,"visible":true,"origin":"","legend":"","description":"","filename":"snvancouveray.bst","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/543cad71e5482964007907cb.bst"},{"id":99312077,"identity":"d43039aa-f6eb-4967-ac24-0af2808f68d3","added_by":"auto","created_at":"2025-12-31 16:17:51","extension":"bst","order_by":36,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40758,"visible":true,"origin":"","legend":"","description":"","filename":"snvancouvernum.bst","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/8c897b4b60b49591f1537fa2.bst"},{"id":99312053,"identity":"393ca9f8-1caa-46b4-b67d-d89fd5ddea19","added_by":"auto","created_at":"2025-12-31 16:17:48","extension":"xml","order_by":37,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":214183,"visible":true,"origin":"","legend":"","description":"","filename":"ad06baad691845be868fa8fcb893b92a1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/1bfdc4fa6e8819f9ccafecc2.xml"},{"id":99311957,"identity":"54355367-12f4-464c-b9a4-7465b26716ea","added_by":"auto","created_at":"2025-12-31 16:17:30","extension":"html","order_by":38,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":188554,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1/f2987bd89d83ed691d72ceb7.html"},{"id":99322998,"identity":"ceb7e5ed-fe80-4fba-b5f9-38e34ce85e23","added_by":"auto","created_at":"2025-12-31 16:44:42","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1400516,"visible":true,"origin":"","legend":"","description":"","filename":"FCLfixed.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8126610/v1_covered_c42acea6-aa92-417f-835b-05d47d65441b.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"FCL: Frequency-based Contrastive Learning for Generalizable Face Forgery Detection","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":"multimedia-systems","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mmsj","sideBox":"Learn more about [Multimedia Systems](http://link.springer.com/journal/530)","snPcode":"530","submissionUrl":"https://submission.nature.com/new-submission/530/3","title":"Multimedia Systems","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Face forgery detection, frequency, contrastive learning, generalization ability","lastPublishedDoi":"10.21203/rs.3.rs-8126610/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8126610/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eWith the proliferation of deepfake techniques, face forgery detection has emerged as a critical area of research to mitigate the risks associated with facial manipulation. Recent studies have highlighted the superiority of frequency information over color-space input, especially in high-compression scenarios. However, most frequency-enhanced methods employ a two-stream network architecture, assigning separate encoders for RGB and frequency domains, and rely solely on vanilla binary cross-entropy loss, limiting their generalization ability.To address these limitations, we propose a novel framework, Frequency-enhanced Contrastive Learning (FCL), which trains the model in a supervised manner using contrastive loss. FCL treats two views of different modalities generated from the same image as positive pairs and samples with opposite labels as negative pairs. This approach pulls features from the RGB and high-frequency domains closer while pushing features of pristine and forgery faces apart. Additionally, we introduce a Shallow Feature Supplement (SFS) module to complement local information from low-level shallow feature embeddings into high-level feature maps, and a Dual Modal Fusion (DMF) module to adaptively aggregate information from both domains.Extensive experiments on seven datasets demonstrate the superior generalization of our method compared to state-of-the-art competitors. Notably, FCL achieves a remarkable accuracy of 92.29% and an AUC of 94.25% on the heavily compressed FaceForensics++ dataset, showcasing its robustness against compression. These results underscore the potential of FCL as a powerful tool for face forgery detection in real-world scenarios.The source code ofour proposed algorithm are available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/Y30230924Wang/Frequency-Enhanced-Contrastive-Learning\u003c/span\u003e\u003cspan address=\"https://github.com/Y30230924Wang/Frequency-Enhanced-Contrastive-Learning\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e","manuscriptTitle":"FCL: Frequency-based Contrastive Learning for Generalizable Face Forgery Detection","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-25 07:27:49","doi":"10.21203/rs.3.rs-8126610/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-02-07T02:54:27+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-02T12:23:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"248532674344625437222979594516703618255","date":"2025-12-30T17:08:25+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-24T04:53:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"91471883794028946540922472180081719255","date":"2025-12-24T00:50:14+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-12-24T00:45:48+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-12-17T03:57:44+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-18T11:28:42+00:00","index":"","fulltext":""},{"type":"submitted","content":"Multimedia Systems","date":"2025-11-16T10:14:52+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"multimedia-systems","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mmsj","sideBox":"Learn more about [Multimedia Systems](http://link.springer.com/journal/530)","snPcode":"530","submissionUrl":"https://submission.nature.com/new-submission/530/3","title":"Multimedia Systems","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"fcb172a2-b710-4e4b-8c2c-85a08b91cd97","owner":[],"postedDate":"December 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-02-27T08:24:32+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-25 07:27:49","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8126610","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8126610","identity":"rs-8126610","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}