A feature enhancement and attention fusion network for small object detection in UAV imagery

A feature enhancement and attention fusion network for small object detection in UAV imagery | 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 A feature enhancement and attention fusion network for small object detection in UAV imagery Xilong Xu, Peng Li, Hongwei Ding, Jinhua Yang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9132561/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Unmanned aerial vehicle (UAV) small object detection suffers from issues like small target scale, complex backgrounds, and motion blur, which severely limit detection accuracy and robustness. To address these challenges, we propose an enhanced detection framework based on YOLOv11. The framework integrates three core modules: a Multi-Perception Feature Fusion (MPFF) module for fine-grained local feature extraction, an Amplitude-Aware Linear Attention (MALA) module for efficient global context modeling, and a Stepwise Attention Fusion (SAF) module for harmonizing local details and global context. A dedicated dynamic detection head is also added to boost small target sensitivity. Experiments on the VisDrone dataset show that our model achieves a 6.24% improvement in mAP50, 4.45% in mAP95, and 4.55% in recall rate, outperforming mainstream detectors while maintaining low computational complexity. Evaluations on the CCTSDB-2021 and RSOD datasets further validate its strong generalization capability. The proposed method provides a practical solution for UAV applications such as urban surveillance, agricultural monitoring, and disaster response. YOLOv11s Multi-Perceptual Feature Fusion Amplitude-aware linear attention Stepwise Attention Fusion Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers invited by journal 19 Mar, 2026 Editor assigned by journal 17 Mar, 2026 Submission checks completed at journal 17 Mar, 2026 First submitted to journal 15 Mar, 2026 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-9132561","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":608695225,"identity":"267af61b-5584-41d8-a44f-4a2dafce44da","order_by":0,"name":"Xilong Xu","email":"","orcid":"","institution":"Yunnan University","correspondingAuthor":false,"prefix":"","firstName":"Xilong","middleName":"","lastName":"Xu","suffix":""},{"id":608695228,"identity":"3dc37719-8df6-4cb8-ac63-de7c53f276a7","order_by":1,"name":"Peng Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxElEQVRIie3RMQrCMBSA4YRAu0Rd00U8wpNCi5CbuLRLNvcOBQOFduzqMQQvUHjgVPAKOUJHBwcjiGve6JCfLIF8kJcwFov9Yym3bAG13QiBjkYEt/zS6DwbEgNE4pecTX19yJ0iiVIIy1Y98htKBqzVxyA5dH6WrEdR4Gpy7G5ONkQAPdn3mBS4roBbJJK6R5l3EhSdTLNRIKjkMwu3jQaF/pEryixlOjjxAnUeR0S3tJpwMf//z9+uCh3/klgsFosFegNFVzul8rEUhgAAAABJRU5ErkJggg==","orcid":"","institution":"Yunnan University","correspondingAuthor":true,"prefix":"","firstName":"Peng","middleName":"","lastName":"Li","suffix":""},{"id":608695234,"identity":"08e2ccc6-ccef-449c-8c7d-45ac8fcb90c4","order_by":2,"name":"Hongwei Ding","email":"","orcid":"","institution":"Dianchi College","correspondingAuthor":false,"prefix":"","firstName":"Hongwei","middleName":"","lastName":"Ding","suffix":""},{"id":608695236,"identity":"008deab3-6f1e-4840-9a90-f959adb43070","order_by":3,"name":"Jinhua Yang","email":"","orcid":"","institution":"Dianchi College","correspondingAuthor":false,"prefix":"","firstName":"Jinhua","middleName":"","lastName":"Yang","suffix":""}],"badges":[],"createdAt":"2026-03-16 03:38:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9132561/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9132561/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105564057,"identity":"c6109b92-9958-4e0a-8938-137039436972","added_by":"auto","created_at":"2026-03-27 12:48:39","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5627872,"visible":true,"origin":"","legend":"","description":"","filename":"AfeatureenhancementandattentionfusionnetworkforsmallobjectdetectioninUAVimagery.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9132561/v1_covered_905637a0-e801-4984-ad49-2febad490b11.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A feature enhancement and attention fusion network for small object detection in UAV imagery","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"signal-image-and-video-processing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"sivp","sideBox":"Learn more about [Signal, Image and Video Processing](http://link.springer.com/journal/11760)","snPcode":"11760","submissionUrl":"https://submission.nature.com/new-submission/11760/3","title":"Signal, Image and Video Processing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"YOLOv11s, Multi-Perceptual Feature Fusion, Amplitude-aware linear attention, Stepwise Attention Fusion","lastPublishedDoi":"10.21203/rs.3.rs-9132561/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9132561/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Unmanned aerial vehicle (UAV) small object detection suffers from issues like small target scale, complex backgrounds, and motion blur, which severely limit detection accuracy and robustness. To address these challenges, we propose an enhanced detection framework based on YOLOv11. The framework integrates three core modules: a Multi-Perception Feature Fusion (MPFF) module for fine-grained local feature extraction, an Amplitude-Aware Linear Attention (MALA) module for efficient global context modeling, and a Stepwise Attention Fusion (SAF) module for harmonizing local details and global context. A dedicated dynamic detection head is also added to boost small target sensitivity. Experiments on the VisDrone dataset show that our model achieves a 6.24% improvement in mAP50, 4.45% in mAP95, and 4.55% in recall rate, outperforming mainstream detectors while maintaining low computational complexity. Evaluations on the CCTSDB-2021 and RSOD datasets further validate its strong generalization capability. The proposed method provides a practical solution for UAV applications such as urban surveillance, agricultural monitoring, and disaster response.","manuscriptTitle":"A feature enhancement and attention fusion network for small object detection in UAV imagery","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-23 03:13:15","doi":"10.21203/rs.3.rs-9132561/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewersInvited","content":"","date":"2026-03-19T07:11:46+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-17T09:31:43+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-17T09:30:53+00:00","index":"","fulltext":""},{"type":"submitted","content":"Signal, Image and Video Processing","date":"2026-03-16T03:23:19+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"signal-image-and-video-processing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"sivp","sideBox":"Learn more about [Signal, Image and Video Processing](http://link.springer.com/journal/11760)","snPcode":"11760","submissionUrl":"https://submission.nature.com/new-submission/11760/3","title":"Signal, Image and Video Processing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"6e9508c1-30f3-4785-bfda-fb6a67abb055","owner":[],"postedDate":"March 23rd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-03-23T03:13:15+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-23 03:13:15","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9132561","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9132561","identity":"rs-9132561","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}