Granularity-Guided Fusion for Multi-Modal Sentiment Understanding

Granularity-Guided Fusion for Multi-Modal Sentiment Understanding | 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 Granularity-Guided Fusion for Multi-Modal Sentiment Understanding Mingxuan Chen, Huarong Tang, Chen Sun, Nannan Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7892962/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 12 Mar, 2026 Read the published version in Scientific Reports → Version 1 posted 11 You are reading this latest preprint version Abstract Multimodal sarcasm detection involves identifying sarcasm through multiple modalities of information, with the key challenge lying in modeling incongruity within and between modalities. Current methods often focus on intermodal incongruity while neglecting the potential of fully exploring semantic information within each modality. To address this, we propose the Granularity-Based Inter and Intra-Modal Fusion Network (GIIFN). This approach integrates handcrafted image descriptors with deep learning models to extract comprehensive semantic information from images and leverages a pre-trained language model to enhance image analysis with large-scale textual knowledge. Moreover, our feature interaction model effectively fuses features at different granularities, capturing fine details and contextual information. Extensive experiments demonstrate that our method outperforms existing approaches and achieves state-of-the-art results in multimodal sarcasm detection. Physical sciences/Engineering Physical sciences/Mathematics and computing Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 12 Mar, 2026 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 08 Dec, 2025 Reviews received at journal 07 Dec, 2025 Reviewers agreed at journal 02 Dec, 2025 Reviews received at journal 30 Nov, 2025 Reviewers agreed at journal 28 Nov, 2025 Reviewers agreed at journal 28 Nov, 2025 Reviewers invited by journal 28 Nov, 2025 Editor assigned by journal 28 Nov, 2025 Editor invited by journal 18 Nov, 2025 Submission checks completed at journal 01 Nov, 2025 First submitted to journal 01 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-7892962","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":552674756,"identity":"62a9a134-7ade-4f70-b7bb-8aacc7db147a","order_by":0,"name":"Mingxuan Chen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAuklEQVRIiWNgGAWjYFACHiCuQGITqeUMyVoY20jRott+9uDjwnl1if38BxgfvG1jkDcnpMXsTF6y8cxthxNnzkhgNpzbxmC4s4GQlhs8ZtK82w4kbrjBwCbN28aQYHCAsBbz37xz6hL3nz/A/ptYLWbMvA3MiRsYEtiYidNyJsdYmufYYeMZNxKbJeeckzDcQFDL8TOGn3lq6mT7+w8f/PCmzEaeoC1IgLEBSEgQr34UjIJRMApGAW4AAMwdPRijyY9kAAAAAElFTkSuQmCC","orcid":"","institution":"Shanghai University of Engineering Science","correspondingAuthor":true,"prefix":"","firstName":"Mingxuan","middleName":"","lastName":"Chen","suffix":""},{"id":552674757,"identity":"437583b1-a9e9-4a5b-8a20-41f7da7bf560","order_by":1,"name":"Huarong Tang","email":"","orcid":"","institution":"Shanghai University of Engineering Science","correspondingAuthor":false,"prefix":"","firstName":"Huarong","middleName":"","lastName":"Tang","suffix":""},{"id":552674758,"identity":"b06d52aa-35c4-476d-a6e9-804f37380ab9","order_by":2,"name":"Chen Sun","email":"","orcid":"","institution":"Shanghai University of Engineering Science","correspondingAuthor":false,"prefix":"","firstName":"Chen","middleName":"","lastName":"Sun","suffix":""},{"id":552674759,"identity":"eb35e872-0592-4abf-9278-48675792e677","order_by":3,"name":"Nannan Zhang","email":"","orcid":"","institution":"Shanghai University of Engineering Science","correspondingAuthor":false,"prefix":"","firstName":"Nannan","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2025-10-18 10:53:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7892962/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7892962/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-026-43363-5","type":"published","date":"2026-03-12T16:00:13+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":97177361,"identity":"2ac61f8d-d4f1-418f-8faa-e89da75558d2","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"json","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5302,"visible":true,"origin":"","legend":"","description":"","filename":"fc2290a6abc64db4be9609365291e3f0.json","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/361cdb751f45b8884a64b9bb.json"},{"id":97177360,"identity":"4f235b46-493b-49c6-b7bb-cbd5a5a7255e","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"png","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":143553,"visible":true,"origin":"","legend":"","description":"","filename":"GranularityLevelAttention.png","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/9fb7088ef17638a113bd0dd3.png"},{"id":97177359,"identity":"556ed191-f78c-4318-8951-8502115be437","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":51899,"visible":true,"origin":"","legend":"","description":"","filename":"SURF.png","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/b92a5961fc6d16b412fcd0b7.png"},{"id":97177364,"identity":"2d3d8876-dad1-45b8-89b8-bc7c910a235e","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":163134,"visible":true,"origin":"","legend":"","description":"","filename":"example.png","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/d93b390779b519ab91d0d00b.png"},{"id":97250853,"identity":"2583070c-31d1-4c10-8578-16b8d1b9612e","added_by":"auto","created_at":"2025-12-02 13:15:27","extension":"ldf","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":49107,"visible":true,"origin":"","legend":"","description":"","filename":"jabbrvltwaall.ldf","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/2af65e3631211089268e5f18.ldf"},{"id":97177366,"identity":"ee6e4eea-8797-4a89-93a6-e9dbaeff179a","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"ldf","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":268657,"visible":true,"origin":"","legend":"","description":"","filename":"jabbrvltwaen.ldf","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/af497c46332980a23f59dc76.ldf"},{"id":97249984,"identity":"1d203047-9af7-4e29-bc3e-4535645a99f0","added_by":"auto","created_at":"2025-12-02 13:13:44","extension":"sty","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15480,"visible":true,"origin":"","legend":"","description":"","filename":"jabbrv.sty","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/0402a14e30258d1b4b3eb613.sty"},{"id":97177362,"identity":"a5bb0395-ea7d-4859-9dc5-b314adbe3e82","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":58270,"visible":true,"origin":"","legend":"","description":"","filename":"knowledge.png","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/2f28a21d7fb657e0b66c8eb2.png"},{"id":97249433,"identity":"6864ec46-d1fd-4a9c-8582-06a8c7a970eb","added_by":"auto","created_at":"2025-12-02 13:12:34","extension":"pdf","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":836987,"visible":true,"origin":"","legend":"","description":"","filename":"main.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/daa6fd0a0c67981043144c8f.pdf"},{"id":97177367,"identity":"54d53f2d-ec5f-4bb9-a7bd-b6722cf7ec79","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":239492,"visible":true,"origin":"","legend":"","description":"","filename":"model.png","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/9193ced175fca7df426725e9.png"},{"id":97177368,"identity":"d1f7082b-9ba5-4c63-9ff7-453e67c9a388","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"bst","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36153,"visible":true,"origin":"","legend":"","description":"","filename":"naturemagdoi.bst","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/a4850dd030c7f3844fc200c7.bst"},{"id":97177371,"identity":"0a931773-74a6-4cec-8368-7319b7acea3d","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"pdf","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":28642,"visible":true,"origin":"","legend":"","description":"","filename":"pointbypointresponse.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/c4b0b6d94a45e8b3f822e5ab.pdf"},{"id":97249182,"identity":"a700d346-33a6-4cb7-a820-9c172dac228a","added_by":"auto","created_at":"2025-12-02 13:11:05","extension":"jpg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":379752,"visible":true,"origin":"","legend":"","description":"","filename":"stream.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/f9adc9ce95424b0032429740.jpg"},{"id":97177369,"identity":"5ca18c1a-de7c-46df-a9b9-b83d2ec6fef7","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"cls","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5824,"visible":true,"origin":"","legend":"","description":"","filename":"wlscirep.cls","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/3bb9f15c1c8721fe34d5152d.cls"},{"id":97177373,"identity":"3196fbbc-e42c-44c7-9d7a-229966e89eab","added_by":"auto","created_at":"2025-12-01 15:49:52","extension":"xml","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":88689,"visible":true,"origin":"","legend":"","description":"","filename":"fc2290a6abc64db4be9609365291e3f01structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1/d07dbdc72efb6d59d68d2300.xml"},{"id":104739634,"identity":"925dc041-bfed-4cd1-8808-0c948facef92","added_by":"auto","created_at":"2026-03-16 16:11:01","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":663596,"visible":true,"origin":"","legend":"","description":"","filename":"main.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7892962/v1_covered_94d1cfbd-2be7-460b-b9f4-c325e6b6f7b8.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Granularity-Guided Fusion for Multi-Modal Sentiment Understanding","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":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7892962/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7892962/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Multimodal sarcasm detection involves identifying sarcasm through multiple modalities of information, with the key challenge lying in modeling incongruity within and between modalities. Current methods often focus on intermodal incongruity while neglecting the potential of fully exploring semantic information within each modality. To address this, we propose the Granularity-Based Inter and Intra-Modal Fusion Network (GIIFN). This approach integrates handcrafted image descriptors with deep learning models to extract comprehensive semantic information from images and leverages a pre-trained language model to enhance image analysis with large-scale textual knowledge. Moreover, our feature interaction model effectively fuses features at different granularities, capturing fine details and contextual information. Extensive experiments demonstrate that our method outperforms existing approaches and achieves state-of-the-art results in multimodal sarcasm detection.","manuscriptTitle":"Granularity-Guided Fusion for Multi-Modal Sentiment Understanding","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-01 15:49:47","doi":"10.21203/rs.3.rs-7892962/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-12-08T06:45:31+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-07T17:29:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"72940192677757367527780587692873141318","date":"2025-12-02T05:26:11+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-01T03:49:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"74074959052836036352907262411562944884","date":"2025-11-28T15:50:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"203079799444610056910800946281665201674","date":"2025-11-28T15:38:15+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-11-28T15:33:10+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-11-28T15:28:21+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-11-18T08:36:46+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-01T06:13:24+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-11-01T06:10:53+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f6207a67-2ad1-49a1-b344-0ace2abbb9bf","owner":[],"postedDate":"December 1st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":58817236,"name":"Physical sciences/Engineering"},{"id":58817237,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2026-03-16T16:06:07+00:00","versionOfRecord":{"articleIdentity":"rs-7892962","link":"https://doi.org/10.1038/s41598-026-43363-5","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2026-03-12 16:00:13","publishedOnDateReadable":"March 12th, 2026"},"versionCreatedAt":"2025-12-01 15:49:47","video":"","vorDoi":"10.1038/s41598-026-43363-5","vorDoiUrl":"https://doi.org/10.1038/s41598-026-43363-5","workflowStages":[]},"version":"v1","identity":"rs-7892962","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7892962","identity":"rs-7892962","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}