HAST: A New Style Transfer Network Integrating Convolution and Attention Mechanism

HAST: A New Style Transfer Network Integrating Convolution and Attention Mechanism | 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 HAST: A New Style Transfer Network Integrating Convolution and Attention Mechanism Kunyun Wu, Yang Xu, Bin Cao, Caideng Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7484474/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 Style transfer is a computer vision technique that aims to apply the artistic style of one image to the content of another, creating a new image that preserves the content of the original image while incorporating the desired artistic style. However, existing style transfer networks still face issues with unclear semantic representation and insufficient detail preservation in stylized images. To address these problems, this paper proposes a novel style transfer network called HAST, which combines convolution and attention mechanisms. Convolution operations help preserve the detailed features, content structure, and semantic information of the image, while the attention mechanism allows the network to focus on important regions or features during image processing, resulting in stylized images with clear details and complete semantics. In the HAST model, the CPCA attention is first improved, enabling the enhanced attention to better focus on image details and adaptively adjust weights according to the network's requirements. Additionally, the image feature extractor SCCA module is designed by combining Parc convolution and the improved CPCA attention, which fully extracts semantic information from the content image and style features from the style image, preparing for subsequent feature fusion. Experimental results show that, with the above design, the HAST network generates images that not only achieve better stylization but also retain clear content semantics, yielding excellent results for arbitrary style transfer. Arbitrary style transfer Improved CPCA attention SCCA module SCCA module Parc convolution 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-7484474","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":510467835,"identity":"07de3512-da0a-454b-a5fe-b115420d558b","order_by":0,"name":"Kunyun Wu","email":"","orcid":"","institution":"Guizhou University","correspondingAuthor":false,"prefix":"","firstName":"Kunyun","middleName":"","lastName":"Wu","suffix":""},{"id":510467836,"identity":"9d1d13dc-c887-4f9f-8af7-2df400042274","order_by":1,"name":"Yang Xu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAr0lEQVRIiWNgGAWjYHACxgcQOoF4LcwGJGthkyBNi8GNHLPKL38OM/Cz5xgw/NxBhBbJGTlmt2V4DjNI9rwxYOw9Q4QWfgmgFgmJwyDrDJgZ24jQwgbUUixhcJjBnmgtIFsYPyQAbZEgVotkz7NiaYYD6TwSZ54VHOwlRovB8eSNH3/8sZbjb0/e+OAnMVoYGDgMmHkYGHhAzANEaWBgYH/A+INIpaNgFIyCUTBCAQBxMTDqBk8JgQAAAABJRU5ErkJggg==","orcid":"","institution":"Guizhou University","correspondingAuthor":true,"prefix":"","firstName":"Yang","middleName":"","lastName":"Xu","suffix":""},{"id":510467837,"identity":"c31bf639-55b9-4e0b-ba4e-46fc6139d45d","order_by":2,"name":"Bin Cao","email":"","orcid":"","institution":"Guizhou University","correspondingAuthor":false,"prefix":"","firstName":"Bin","middleName":"","lastName":"Cao","suffix":""},{"id":510467838,"identity":"42c30e37-25c3-4086-9cb0-2bf6676ca87d","order_by":3,"name":"Caideng Zhang","email":"","orcid":"","institution":"Guizhou University","correspondingAuthor":false,"prefix":"","firstName":"Caideng","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2025-08-29 03:38:02","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7484474/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7484474/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103995988,"identity":"a0418c99-2239-44f7-91e7-9120dd57b4ce","added_by":"auto","created_at":"2026-03-05 12:41:59","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1923979,"visible":true,"origin":"","legend":"","description":"","filename":"HAST2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7484474/v1_covered_f3ed610e-ae82-42a4-8e3b-6f528d49683a.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"HAST: A New Style Transfer Network Integrating Convolution and Attention Mechanism","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":"Arbitrary style transfer, Improved CPCA attention, SCCA module, SCCA module, Parc convolution","lastPublishedDoi":"10.21203/rs.3.rs-7484474/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7484474/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eStyle transfer is a computer vision technique that aims to apply the artistic style of one image to the content of another, creating a new image that preserves the content of the original image while incorporating the desired artistic style. However, existing style transfer networks still face issues with unclear semantic representation and insufficient detail preservation in stylized images. To address these problems, this paper proposes a novel style transfer network called HAST, which combines convolution and attention mechanisms. Convolution operations help preserve the detailed features, content structure, and semantic information of the image, while the attention mechanism allows the network to focus on important regions or features during image processing, resulting in stylized images with clear details and complete semantics. In the HAST model, the CPCA attention is first improved, enabling the enhanced attention to better focus on image details and adaptively adjust weights according to the network's requirements. Additionally, the image feature extractor SCCA module is designed by combining Parc convolution and the improved CPCA attention, which fully extracts semantic information from the content image and style features from the style image, preparing for subsequent feature fusion. Experimental results show that, with the above design, the HAST network generates images that not only achieve better stylization but also retain clear content semantics, yielding excellent results for arbitrary style transfer.\u003c/p\u003e","manuscriptTitle":"HAST: A New Style Transfer Network Integrating Convolution and Attention Mechanism","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-11 07:34:00","doi":"10.21203/rs.3.rs-7484474/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":"d57a8fd4-3724-44e0-8adf-6897c74f722b","owner":[],"postedDate":"September 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-05T12:41:21+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-11 07:34:00","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7484474","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7484474","identity":"rs-7484474","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}