Global Vision, Local Focus: The Semantic Enhancement Transformer Network for Crowd Counting

Global Vision, Local Focus: The Semantic Enhancement Transformer Network for Crowd Counting | 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 Global Vision, Local Focus: The Semantic Enhancement Transformer Network for Crowd Counting Mingtao Wang, Xin Zhou, Yuanyuan Chen This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3938792/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Feb, 2025 Read the published version in Soft Computing → Version 1 posted 5 You are reading this latest preprint version Abstract Automatic crowd counting has made significant progress in recent years. However, due to the challenge of multi-scale variations, convolutional neural networks (CNNs) with fixed-size kernels cannot effectively address this difficulty, leading to a severe limitation on counting performance. To alleviate this issue, we propose a semantic enhancement Transformer crowd counting network (named SET) to improve the semantic encoding relationships in crowd scenes. The SET integrates global attention from Transformer, learnable local attention, and inductive bias from CNNs into a counting model. Firstly, we introduce an efficient Transformer encoder to extract low-level global features of crowd scenes. Secondly, we propose a learnable ViTBlock to dynamically learn appropriate weights for different regions, aiding in enhancing the model’s global visual understanding. Finally, to guide the model to focus better on crowd regions, we jointly employ a segmentation attention module and a feature aggregation module to aggregate semantic and spatial features at multiple levels, thus obtaining finer-grained features. We conduct extensive experiments on four challenging datasets, including ShanghaiTech Part A/B, UCF-QNRF, and JHU-CROWD++, achieving excellent results. Crowd counting Semantic enhancement Segmentation attention Feature aggregation Full Text Cite Share Download PDF Status: Published Journal Publication published 07 Feb, 2025 Read the published version in Soft Computing → Version 1 posted Editorial decision: Major Revision 06 May, 2024 Reviewers agreed at journal 19 Apr, 2024 Reviewers invited by journal 01 Apr, 2024 Editor invited by journal 27 Mar, 2024 First submitted to journal 06 Feb, 2024 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-3938792","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":286309161,"identity":"d02483ae-5423-4055-a749-578dc869f60b","order_by":0,"name":"Mingtao Wang","email":"","orcid":"","institution":"Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Mingtao","middleName":"","lastName":"Wang","suffix":""},{"id":286309162,"identity":"8cb6ab2b-750a-4633-a38f-25e31ca1011d","order_by":1,"name":"Xin Zhou","email":"","orcid":"","institution":"Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Zhou","suffix":""},{"id":286309163,"identity":"b8b28345-675f-4eb6-9a95-9ad93a919d79","order_by":2,"name":"Yuanyuan Chen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA80lEQVRIiWNgGAWjYDACCWTOBwhlQLwWxhkka2HmIUYL/+zmY4952+zyDG4kP3ts88cusYG9eZsEQ80d3JbcOZZuzNuWXGxwI83cOLctObGB51iZBMOxZzi1GEjkmEnztjEnbriRYCad23AgsQEoIsHYcBiPlvxvQC31QC3p36Qt/gC1yL8hpCWHDajlMFAL0DoGNpAtPPi1SNxIM5Occ+544swzb8oke9uSjdt40ootEo7h1sI/I/mZxJuy6sS+4+nbJH78sZPtZz+88caHGtxaQIAJFB0KB6A8NhCRgFcDMNJ/AAn5BgKqRsEoGAWjYOQCAKTOU7WehAQPAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-5358-9213","institution":"Sichuan University","correspondingAuthor":true,"prefix":"","firstName":"Yuanyuan","middleName":"","lastName":"Chen","suffix":""}],"badges":[],"createdAt":"2024-02-08 04:38:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3938792/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3938792/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00500-025-10506-1","type":"published","date":"2025-02-07T15:58:22+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":75930793,"identity":"effcd8f5-db73-456c-9a43-1199f85b5f39","added_by":"auto","created_at":"2025-02-10 16:13:25","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":11854855,"visible":true,"origin":"","legend":"","description":"","filename":"CrowdEST.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3938792/v1_covered_6efe77a6-1d17-4338-9abd-4cf456a4448d.pdf"}],"financialInterests":"","formattedTitle":"Global Vision, Local Focus: The Semantic Enhancement Transformer Network for Crowd Counting","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":true,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"soft-computing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"soco","sideBox":"Learn more about [Soft Computing](https://www.springer.com/journal/500)","snPcode":"500","submissionUrl":"https://submission.nature.com/new-submission/500/3","title":"Soft Computing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Crowd counting, Semantic enhancement, Segmentation attention, Feature aggregation","lastPublishedDoi":"10.21203/rs.3.rs-3938792/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3938792/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Automatic crowd counting has made significant progress in recent years. However, due to the challenge of multi-scale variations, convolutional neural networks (CNNs) with fixed-size kernels cannot effectively address this difficulty, leading to a severe limitation on counting performance. To alleviate this issue, we propose a semantic enhancement Transformer crowd counting network (named SET) to improve the semantic encoding relationships in crowd scenes. The SET integrates global attention from Transformer, learnable local attention, and inductive bias from CNNs into a counting model. Firstly, we introduce an efficient Transformer encoder to extract low-level global features of crowd scenes. Secondly, we propose a learnable ViTBlock to dynamically learn appropriate weights for different regions, aiding in enhancing the model’s global visual understanding. Finally, to guide the model to focus better on crowd regions, we jointly employ a segmentation attention module and a feature aggregation module to aggregate semantic and spatial features at multiple levels, thus obtaining finer-grained features. We conduct extensive experiments on four challenging datasets, including ShanghaiTech Part A/B, UCF-QNRF, and JHU-CROWD++, achieving excellent results.","manuscriptTitle":"Global Vision, Local Focus: The Semantic Enhancement Transformer Network for Crowd Counting","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-04 06:45:06","doi":"10.21203/rs.3.rs-3938792/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Major Revision","date":"2024-05-06T04:07:25+00:00","index":"","fulltext":""},{"type":"reviewerAgreed","content":"","date":"2024-04-20T02:11:31+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-04-01T20:03:36+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"Soft Computing","date":"2024-03-27T09:48:25+00:00","index":"","fulltext":""},{"type":"submitted","content":"Soft Computing","date":"2024-02-06T22:22:42+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"soft-computing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"soco","sideBox":"Learn more about [Soft Computing](https://www.springer.com/journal/500)","snPcode":"500","submissionUrl":"https://submission.nature.com/new-submission/500/3","title":"Soft Computing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"1af82b2c-ca5d-407c-a5ae-f0f1b0071653","owner":[],"postedDate":"April 4th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-02-10T16:06:10+00:00","versionOfRecord":{"articleIdentity":"rs-3938792","link":"https://doi.org/10.1007/s00500-025-10506-1","journal":{"identity":"soft-computing","isVorOnly":false,"title":"Soft Computing"},"publishedOn":"2025-02-07 15:58:22","publishedOnDateReadable":"February 7th, 2025"},"versionCreatedAt":"2024-04-04 06:45:06","video":"","vorDoi":"10.1007/s00500-025-10506-1","vorDoiUrl":"https://doi.org/10.1007/s00500-025-10506-1","workflowStages":[]},"version":"v1","identity":"rs-3938792","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3938792","identity":"rs-3938792","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}