Research on an improved small target detection algorithm for YOLOv8

Research on an improved small target detection algorithm for YOLOv8 | 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 an improved small target detection algorithm for YOLOv8 Tinghang GUO, Xiaohan LI, Xin JI, Zuanping QIN, Guangda LU, Yu HAN, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6757420/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 In order to solve the problems of missed detection and low detection accuracy due to the changes in the shape, appearance and position of objects in the imaging process, as well as the complex influence of lighting conditions and occlusion factors, an improved YOLOv8 algorithm based on Swin Transformer was proposed. By introducing modules including Focus, deeply separable convolutional DwConv, c2, etc., the computation and parameters are reduced, the receptive field and feature channels are increased, and the Swin Transformer module is used to extract visual features to capture the context information of small target objects to enhance the feature representation. In addition, the loss function of the original network is replaced by the WIOU loss function to optimize the model and improve the accuracy of small target detection. Comparative experiments based on public datasets show that compared with the YOLOv8 algorithm, the improved model has improved the accuracy of small target detection P, recall R and average accuracy of [email protected] , which enhances the ability of intelligent robots to identify small targets in complex environments and provides important support for the technological progress of related industries. Physical sciences/Physics/Techniques and instrumentation Physical sciences/Mathematics and computing/Computer science small target detection deep learning improved YOLOV8 model Loss 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-6757420","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":486029805,"identity":"9d8dd3f8-53c6-440f-a1b5-4c713093191c","order_by":0,"name":"Tinghang GUO","email":"","orcid":"","institution":"Tianjin University of Technology and Education","correspondingAuthor":false,"prefix":"","firstName":"Tinghang","middleName":"","lastName":"GUO","suffix":""},{"id":486029806,"identity":"a446f281-2981-4bc9-9474-a29ed8148086","order_by":1,"name":"Xiaohan LI","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3ElEQVRIiWNgGAWjYBACfvbm4x8+GPyr55d/fIA4LZI9x9IYZxQcSJBsSEsgTovBjRwzZp4PBxIMDuQYEOmyAzlmj3kM7uQZHDjz8cYbBjs53QYCOhgbjpUbzjF4Vix5sHez5RyGZGOzAwS0MDM2b5B4Y8DM2HeYd5s0D8OBxG2EtLAxMxhI8AC1NBzjeUacFh42FjNJHoPDiRPO8LARp0WChy3ZcIZBmrHkDDZjyzkGRPjF/v7jgw8+/LGR45dgfnjjTYWdHEEtaFYSGzVIWkjVMQpGwSgYBSMCAACi8UaNSX60tAAAAABJRU5ErkJggg==","orcid":"","institution":"Tianjin University of Technology and Education","correspondingAuthor":true,"prefix":"","firstName":"Xiaohan","middleName":"","lastName":"LI","suffix":""},{"id":486029807,"identity":"44b2e6b5-73d1-4cde-b20e-7b87a0ce926d","order_by":2,"name":"Xin JI","email":"","orcid":"","institution":"Tianjin University of Technology and Education","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"JI","suffix":""},{"id":486029808,"identity":"d22f38ed-0495-451d-8f16-15f36890f98d","order_by":3,"name":"Zuanping QIN","email":"","orcid":"","institution":"Tianjin University of Technology and Education","correspondingAuthor":false,"prefix":"","firstName":"Zuanping","middleName":"","lastName":"QIN","suffix":""},{"id":486029809,"identity":"9a33cfa4-9d4c-4c32-af89-bcc1ce30dd09","order_by":4,"name":"Guangda LU","email":"","orcid":"","institution":"Tianjin University of Technology and Education","correspondingAuthor":false,"prefix":"","firstName":"Guangda","middleName":"","lastName":"LU","suffix":""},{"id":486029810,"identity":"dc9839a9-6dfe-4bef-950d-309a7d3d89a0","order_by":5,"name":"Yu HAN","email":"","orcid":"","institution":"Tianjin University of Technology and Education","correspondingAuthor":false,"prefix":"","firstName":"Yu","middleName":"","lastName":"HAN","suffix":""},{"id":486029811,"identity":"a95cafbd-3f7a-4315-b758-edc6eaedf77d","order_by":6,"name":"Runze LI","email":"","orcid":"","institution":"Tianjin University of Technology and Education","correspondingAuthor":false,"prefix":"","firstName":"Runze","middleName":"","lastName":"LI","suffix":""}],"badges":[],"createdAt":"2025-05-27 08:53:35","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6757420/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6757420/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":88424433,"identity":"dc128670-5414-4a93-ba5b-4b66793b4de0","added_by":"auto","created_at":"2025-08-06 09:47:19","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":844984,"visible":true,"origin":"","legend":"","description":"","filename":"ResearchonmultiscenesmalltargetdetectionalgorithmbasedonimprovedYOLOv81.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6757420/v1_covered_067db5d7-856e-4a83-a478-81101cf75cad.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Research on an improved small target detection algorithm for YOLOv8","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":"small target detection, deep learning, improved YOLOV8 model, Loss","lastPublishedDoi":"10.21203/rs.3.rs-6757420/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6757420/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"In order to solve the problems of missed detection and low detection accuracy due to the changes in the shape, appearance and position of objects in the imaging process, as well as the complex influence of lighting conditions and occlusion factors, an improved YOLOv8 algorithm based on Swin Transformer was proposed. By introducing modules including Focus, deeply separable convolutional DwConv, c2, etc., the computation and parameters are reduced, the receptive field and feature channels are increased, and the Swin Transformer module is used to extract visual features to capture the context information of small target objects to enhance the feature representation. In addition, the loss function of the original network is replaced by the WIOU loss function to optimize the model and improve the accuracy of small target detection. Comparative experiments based on public datasets show that compared with the YOLOv8 algorithm, the improved model has improved the accuracy of small target detection P, recall R and average accuracy of [email protected], which enhances the ability of intelligent robots to identify small targets in complex environments and provides important support for the technological progress of related industries.","manuscriptTitle":"Research on an improved small target detection algorithm for YOLOv8","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-17 14:44:47","doi":"10.21203/rs.3.rs-6757420/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":"62f1e946-eaf7-47ab-9d79-b311efcb96d0","owner":[],"postedDate":"July 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":51595880,"name":"Physical sciences/Physics/Techniques and instrumentation"},{"id":51595881,"name":"Physical sciences/Mathematics and computing/Computer science"}],"tags":[],"updatedAt":"2025-08-06T09:39:11+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-17 14:44:47","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6757420","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6757420","identity":"rs-6757420","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}