DRIPNet: Decoupled Region of Interest Pooling Feature Network Based on Diffusion Model for UAV Small Object Detection

DRIPNet: Decoupled Region of Interest Pooling Feature Network Based on Diffusion Model for UAV Small Object 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 DRIPNet: Decoupled Region of Interest Pooling Feature Network Based on Diffusion Model for UAV Small Object Detection Xiao Wang, Haijiang Zhu, Zhiqing Li This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4244827/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 Small object detection has always been one of the most challenging tasks in the computer vision. Up to now, a prior bounding box is often applied to Unmanned Aerial Vehicle (UAV) image object detection. However, anchors need to be pre-set and not optimal for training data in many object detection algorithms. In 2022, the Diffusion Model was introduced in object detection method, in which the random boxes are employed. Inspired by this approach and the characteristics of UAV images, we find the great potential of diffusion models in UAV image detection and propose a more reasonable Decoupled Region of Interest Pooling Feature Diffusion Network. First of all, a more rational decoupled region of interest pooling(DRIP) feature extraction module has been designed, which decouples the feature extraction process between different scales, to make full use of the features at each level of the pyramid. Our approach eliminates the negative effects of unreasonable bounding box assignments, thereby enhancing the overall performance. Secondly, we propose a high-resolution scale-varying robust backbone(HSRB), where we architect the convolution module in the backbone using atrous convolution with switchable atrous rates and Pixel-Shuffle upsampling to mitigate the negative effects of scale variation and downsampling. Finally,loss functions with normalized Wasserstein distance (NWD) terms are applied, NWD is led into measuring the similarity between the prediction box and the ground truth box. The purpose is to eliminate the influence of positional sensitivity on the matching between the predicted box and the ground truth box.The optimal results of 27.91% mAP on the VisDrone dataset and 8.42% mAP on the TinyPerson dataset demonstrate the effectiveness of the proposed model. Diffusion model Small object detection UAV image Switchable atrous convolution Pixel-Shuffle 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-4244827","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":289744170,"identity":"32beb352-6b4e-4a99-9bf6-cc861a93d535","order_by":0,"name":"Xiao Wang","email":"","orcid":"","institution":"Beijing University of Chemical Technology","correspondingAuthor":false,"prefix":"","firstName":"Xiao","middleName":"","lastName":"Wang","suffix":""},{"id":289744171,"identity":"ba5312e6-eb96-4741-a451-0c64647d3474","order_by":1,"name":"Haijiang Zhu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6UlEQVRIiWNgGAWjYDACCQY2BokKBgY2diADAhKI0XIGqIUZrMWASC2MbUAG0VrkZ/eYPbCct02eD6jlMU/NHwZ+9hwDhp87cGthnHPG3EBy223DNmYGdmOeYwYMkj1vDBh7z+DWwiyRYyYB1MII1MImzdtgwGBwI8eAGexUHIANrGXObXu4FntCWnjAWhpuJyJskSCgRUIirUxC4tjt5DZmxnbDOceMeSTOPCs42ItHi/yM5G3SEjW3bee3Nx978KZGTo6/PXnjg594tECCAEwxNoBdCiIO4NcAVPuBkIpRMApGwSgY2QAA0+VAeq95dXYAAAAASUVORK5CYII=","orcid":"","institution":"Beijing University of Chemical Technology","correspondingAuthor":true,"prefix":"","firstName":"Haijiang","middleName":"","lastName":"Zhu","suffix":""},{"id":289744174,"identity":"d00b79d6-f5a7-448d-aeac-3740b59f819a","order_by":2,"name":"Zhiqing Li","email":"","orcid":"","institution":"Beijing University of Chemical Technology","correspondingAuthor":false,"prefix":"","firstName":"Zhiqing","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2024-04-10 03:29:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4244827/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4244827/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":56851959,"identity":"ea0485c2-00f8-4395-8c4f-2e92ad36ed14","added_by":"auto","created_at":"2024-05-21 09:18:47","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2594428,"visible":true,"origin":"","legend":"","description":"","filename":"snarticletemplate.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4244827/v1_covered_b7287656-3eba-489e-a762-0c0270d76d7d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"DRIPNet: Decoupled Region of Interest Pooling Feature Network Based on Diffusion Model for UAV Small Object Detection","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":"Diffusion model, Small object detection,UAV image, Switchable atrous convolution,Pixel-Shuffle","lastPublishedDoi":"10.21203/rs.3.rs-4244827/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4244827/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Small object detection has always been one of the most challenging tasks in the computer vision. Up to now, a prior bounding box is often applied to Unmanned Aerial Vehicle (UAV) image object detection. However, anchors need to be pre-set and not optimal for training data in many object detection algorithms. In 2022, the Diffusion Model was introduced in object detection method, in which the random boxes are employed. Inspired by this approach and the characteristics of UAV images, we find the great potential of diffusion models in UAV image detection and propose a more reasonable Decoupled Region of Interest Pooling Feature Diffusion Network. First of all, a more rational decoupled region of interest pooling(DRIP) feature extraction module has been designed, which decouples the feature extraction process between different scales, to make full use of the features at each level of the pyramid. Our approach eliminates the negative effects of unreasonable bounding box assignments, thereby enhancing the overall performance. Secondly, we propose a high-resolution scale-varying robust backbone(HSRB), where we architect the convolution module in the backbone using atrous convolution with switchable atrous rates and Pixel-Shuffle upsampling to mitigate the negative effects of scale variation and downsampling. Finally,loss functions with normalized Wasserstein distance (NWD) terms are applied, NWD is led into measuring the similarity between the prediction box and the ground truth box. The purpose is to eliminate the influence of positional sensitivity on the matching between the predicted box and the ground truth box.The optimal results of 27.91% mAP on the VisDrone dataset and 8.42% mAP on the TinyPerson dataset demonstrate the effectiveness of the proposed model.","manuscriptTitle":"DRIPNet: Decoupled Region of Interest Pooling Feature Network Based on Diffusion Model for UAV Small Object Detection","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-15 02:33:29","doi":"10.21203/rs.3.rs-4244827/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":"a2c87d9b-2852-4b97-ae0f-4487bfd8130b","owner":[],"postedDate":"April 15th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-05-21T09:10:38+00:00","versionOfRecord":[],"versionCreatedAt":"2024-04-15 02:33:29","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4244827","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4244827","identity":"rs-4244827","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}