Hybrid Sparse and Dense Attentions of Similar Regions for Image Denoising

Hybrid Sparse and Dense Attentions of Similar Regions for Image Denoising | 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 Hybrid Sparse and Dense Attentions of Similar Regions for Image Denoising Daiqiang Chen This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4154630/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 02 Apr, 2025 Read the published version in Signal, Image and Video Processing → Version 1 posted 5 You are reading this latest preprint version Abstract Self-attention based on dot-product has achieved great success in the field of computer vision. Its effectiveness is owed to the large capacity of capturing long-range dependencies in feature maps. However, its quadratic computational complexity with respect to the image size hinders the further application of the self-attention modules. Therefore, a variety of strategies, which limit the regions of the computation of dot-product, have been proposed to reduce the computational amount. Through analyzing the advantages and disadvantages of these methods, we introduce a hybrid sparse and dense attention module (HSDA) which adopts the dense dot-product attention in most similar regions and the sparse attention in other regions. Numerical experiments on image denoising demonstrate that the designed HSDA module has the advantage of both sparse and local dense attentions, and can obtain similar PSNRs to full attention at lower computational amount. The corresponding network constructed by the HSDA modules can product favorable results compared to many state-of-the-art methods. Image denoising Self-attention Sparse attention Dense attention Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 02 Apr, 2025 Read the published version in Signal, Image and Video Processing → Version 1 posted Editorial decision: Revision requested 13 Aug, 2024 Reviewers invited by journal 30 Mar, 2024 Submission checks completed at journal 26 Mar, 2024 Editor assigned by journal 26 Mar, 2024 First submitted to journal 23 Mar, 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-4154630","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":283903863,"identity":"872f2724-7c6d-4b3a-b171-8e1aec8e538e","order_by":0,"name":"Daiqiang Chen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAx0lEQVRIiWNgGAWjYDACCQhZz8be2PjgAwlabBL4eQ43G84gQUtaguSM9DZpDmJ08M9uPvbg457DeQY3HzZIMzDYyek2ELLkzrF0wxnPDhcb3E5sMC5gSDY2O0BAi4FEjpk0z4HDjBuAWpJnMBxI3EZYS/436T8gLTcPNhzmIU5LDps0w4G0xJkzGBubidIicSPNTLLngI0xP09iM+MMAyL8wj8j+ZnEjwMScmzsx5//+FBhJ0dQC7o7SVM+CkbBKBgFowAHAAATeUXzC3gNTwAAAABJRU5ErkJggg==","orcid":"","institution":"Army Medical University","correspondingAuthor":true,"prefix":"","firstName":"Daiqiang","middleName":"","lastName":"Chen","suffix":""}],"badges":[],"createdAt":"2024-03-23 13:39:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4154630/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4154630/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s11760-025-04012-y","type":"published","date":"2025-04-02T15:56:58+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":80082002,"identity":"2681acce-d0bd-475c-9501-909a56c76455","added_by":"auto","created_at":"2025-04-07 16:05:30","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5277579,"visible":true,"origin":"","legend":"","description":"","filename":"HybridSparseandDenseAttentionsofSimilarRegions.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4154630/v1_covered_d13a77e3-8e19-476b-9353-1b38219b2449.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Hybrid Sparse and Dense Attentions of Similar Regions for Image Denoising","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":"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":"Image denoising, Self-attention, Sparse attention, Dense attention","lastPublishedDoi":"10.21203/rs.3.rs-4154630/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4154630/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Self-attention based on dot-product has achieved great success in the field of computer vision. Its effectiveness is owed to the large capacity of capturing long-range dependencies in feature maps. However, its quadratic computational complexity with respect to the image size hinders the further application of the self-attention modules. Therefore, a variety of strategies, which limit the regions of the computation of dot-product, have been proposed to reduce the computational amount. Through analyzing the advantages and disadvantages of these methods, we introduce a hybrid sparse and dense attention module (HSDA) which adopts the dense dot-product attention in most similar regions and the sparse attention in other regions. Numerical experiments on image denoising demonstrate that the designed HSDA module has the advantage of both sparse and local dense attentions, and can obtain similar PSNRs to full attention at lower computational amount. The corresponding network constructed by the HSDA modules can product favorable results compared to many state-of-the-art methods.","manuscriptTitle":"Hybrid Sparse and Dense Attentions of Similar Regions for Image Denoising","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-03-28 03:31:19","doi":"10.21203/rs.3.rs-4154630/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-08-13T23:03:22+00:00","index":"","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-03-30T19:28:34+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-03-26T05:49:39+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-03-26T05:49:39+00:00","index":"","fulltext":""},{"type":"submitted","content":"Signal, Image and Video Processing","date":"2024-03-23T13:37:51+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":"78a1d7e7-87c7-4e73-939f-06ff2c72f81a","owner":[],"postedDate":"March 28th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-04-07T15:59:51+00:00","versionOfRecord":{"articleIdentity":"rs-4154630","link":"https://doi.org/10.1007/s11760-025-04012-y","journal":{"identity":"signal-image-and-video-processing","isVorOnly":false,"title":"Signal, Image and Video Processing"},"publishedOn":"2025-04-02 15:56:58","publishedOnDateReadable":"April 2nd, 2025"},"versionCreatedAt":"2024-03-28 03:31:19","video":"","vorDoi":"10.1007/s11760-025-04012-y","vorDoiUrl":"https://doi.org/10.1007/s11760-025-04012-y","workflowStages":[]},"version":"v1","identity":"rs-4154630","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4154630","identity":"rs-4154630","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}