DFMIR-Net: Dual-Frequency Mamba Network for Single-Image Deraining

DFMIR-Net: Dual-Frequency Mamba Network for Single-Image Deraining | 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 DFMIR-Net: Dual-Frequency Mamba Network for Single-Image Deraining Yun Jiang, Keyuan Zhou, Kunyi Zhu, Xijie Wang, Pengyu Chen This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7996606/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 11 You are reading this latest preprint version Abstract Single-image deraining aims to recover a clean image from inputs degraded by rain streaks, yet remains challenging because the streaks are cross-scale, anisotropic, and spatially complex. Existing approaches share several limitations: prior-driven methods are brittle under complex rain; CNNs, though efficient, are confined to local modeling and lack long-range dependencies; Transformers capture global context but have O(N²) time and memory complexity with global self-attention and may oversmooth details; state-space models (SSMs) offer near-linear complexity but often underemphasize high-frequency details; and frequency-only designs risk ringing artifacts and unstable spatial–spectral fusion. To address these issues, we combine global dependency modeling, lightweight local texture enhancement, and an explicit frequency prior under near-linear complexity, balancing global consistency with fine-detail fidelity. Building on this idea, we propose DFMIR-Net: selective SSMs provide efficient long-range modeling to complement CNNs and Transformers; a lightweight local enhancement mechanism strengthens textures and edges, improving performance in dense rain; and a rain-aware frequency module at the bottleneck employs a phase-preserving spectral representation with adaptive fusion so that spatial features and high-frequency cues jointly constrain restoration, mitigating SSMs’ high-frequency underemphasis and stabilizing spatial–spectral fusion. Experiments on Rain100L/H and Test100/1200/2800 show that DFMIR-Net achieves competitive or improved PSNR and SSIM compared with CNN-, Transformer-, and SSM-based methods, yields clearer visual details, and maintains near-linear computation and memory scaling. Single-image deraining Mamba State-space models Frequency-domain learning Image restoration Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 31 Jan, 2026 Reviews received at journal 30 Jan, 2026 Reviews received at journal 13 Jan, 2026 Reviewers agreed at journal 12 Jan, 2026 Reviewers agreed at journal 09 Jan, 2026 Reviewers agreed at journal 08 Jan, 2026 Reviewers agreed at journal 08 Jan, 2026 Reviewers invited by journal 06 Jan, 2026 Editor assigned by journal 06 Jan, 2026 Submission checks completed at journal 03 Nov, 2025 First submitted to journal 31 Oct, 2025 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-7996606","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":571056402,"identity":"207ddb38-1e4e-40f8-bad2-47e7ea0dc38c","order_by":0,"name":"Yun Jiang","email":"","orcid":"","institution":"Northwest Normal University","correspondingAuthor":false,"prefix":"","firstName":"Yun","middleName":"","lastName":"Jiang","suffix":""},{"id":571056403,"identity":"c1a30055-f6a5-4c26-9872-680266c797e3","order_by":1,"name":"Keyuan Zhou","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABEUlEQVRIie3Ov0vDQBTA8TsO0uUg6yuF/g0vHIhFbP+VOwI6NShZOklC4LpU5wj+ER0drxzokj8g0MEWBxeHgCAVChp/bJqQ0eG+wxuO9+EdIS7XP8xnzGCFMPR7abqpfl6hjfTnWp7nZ4eiv7BZkHchWBT4wquZwvJUD3gXQkoZLDmCJEZl4vjWRmjYas3JOGoSNJchAkJEk1X6OC1sjMYLjzgJ4ybCQN4hIsSMppmY6rVaGn5Q/9CopIF4oOY7iaA0o3ow+iL+ayvh3BI0NVl4NaHfV7xWAj1NggRBAKdZcKnf1bX1xOgGw0YysX4l9vuL4eThabt50yfq6j7bls+zcSP5I/Y5sPu+y+VyuX73ARg4XPFRVy/8AAAAAElFTkSuQmCC","orcid":"","institution":"Northwest Normal University","correspondingAuthor":true,"prefix":"","firstName":"Keyuan","middleName":"","lastName":"Zhou","suffix":""},{"id":571056404,"identity":"818f5b7c-4961-47d1-b3bf-c888b15c403a","order_by":2,"name":"Kunyi Zhu","email":"","orcid":"","institution":"Northwest Normal University","correspondingAuthor":false,"prefix":"","firstName":"Kunyi","middleName":"","lastName":"Zhu","suffix":""},{"id":571056405,"identity":"00ab76ea-5db3-4c67-bae4-6be37462e138","order_by":3,"name":"Xijie Wang","email":"","orcid":"","institution":"Northwest Normal University","correspondingAuthor":false,"prefix":"","firstName":"Xijie","middleName":"","lastName":"Wang","suffix":""},{"id":571056406,"identity":"ace14ef2-c4d8-47ab-9521-e3b837418215","order_by":4,"name":"Pengyu Chen","email":"","orcid":"","institution":"Northwest Normal University","correspondingAuthor":false,"prefix":"","firstName":"Pengyu","middleName":"","lastName":"Chen","suffix":""}],"badges":[],"createdAt":"2025-10-31 09:08:21","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7996606/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7996606/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":100356903,"identity":"d113502d-d7fd-4331-84ef-9966d7703505","added_by":"auto","created_at":"2026-01-16 07:17:56","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8319160,"visible":true,"origin":"","legend":"","description":"","filename":"dfmirnet.docx","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/71e561dee394f085251fa33f.docx"},{"id":99830972,"identity":"844588a8-e761-4a9b-8e04-fc954bed5de2","added_by":"auto","created_at":"2026-01-08 17:32:13","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7297,"visible":true,"origin":"","legend":"","description":"","filename":"9d42161d153840fdb6b038ce5981dea1.json","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/fb85f36ea7764c3db97111e8.json"},{"id":99830981,"identity":"f6725faf-20c6-422a-9686-4c4e40adb77a","added_by":"auto","created_at":"2026-01-08 17:32:13","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":147520,"visible":true,"origin":"","legend":"","description":"","filename":"9d42161d153840fdb6b038ce5981dea11enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/8df111fbf41a217e5d1c29b8.xml"},{"id":100356500,"identity":"0a27f678-b346-4e75-a466-f37383ca1ee0","added_by":"auto","created_at":"2026-01-16 07:13:45","extension":"jpeg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":345385,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/e29dfd72795b81b79740c4fa.jpeg"},{"id":99830974,"identity":"07910f94-fd00-4120-a180-dc5746891949","added_by":"auto","created_at":"2026-01-08 17:32:13","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":71175,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/69c396f8833b0dcf3ae68bb5.jpeg"},{"id":100356556,"identity":"c9d067a0-3f81-4831-a01b-b29027a2ff8c","added_by":"auto","created_at":"2026-01-16 07:15:31","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":52386,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/31405bdc16fecea7d0830a31.jpeg"},{"id":100356342,"identity":"083ca7af-349d-4f7a-9dbd-06f611311d5a","added_by":"auto","created_at":"2026-01-16 07:03:52","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":223727,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/7329ca21dc38703e099df9de.jpeg"},{"id":100356997,"identity":"7a52e836-ac36-4da0-a139-bc5f768266ef","added_by":"auto","created_at":"2026-01-16 07:18:14","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":76987,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/394cce51965405836dbd6714.jpeg"},{"id":100356689,"identity":"11609e91-c32c-4a3c-bfa1-e35ed956770c","added_by":"auto","created_at":"2026-01-16 07:16:59","extension":"emf","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1810012,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.emf","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/830aacf63526e563a5a641cf.emf"},{"id":100356351,"identity":"7afd4168-3f23-4f61-af17-d0cfe09bba64","added_by":"auto","created_at":"2026-01-16 07:04:19","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1266336,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/4be56728bdb15ebe04da2731.jpeg"},{"id":99830979,"identity":"1ea9627c-9255-47fe-9710-e215cc909e44","added_by":"auto","created_at":"2026-01-08 17:32:13","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":66721,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/bbaae7d0b070dd276d36a645.png"},{"id":99830978,"identity":"12598b53-997f-45e1-928f-9db76d086ee6","added_by":"auto","created_at":"2026-01-08 17:32:13","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12049,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/87a0da60268324ca1ae7001a.png"},{"id":99830984,"identity":"24aa8ee5-2176-456c-81e1-38c2ab773285","added_by":"auto","created_at":"2026-01-08 17:32:13","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":10549,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/d03e79737df43fed79432608.png"},{"id":99830986,"identity":"731ddb4f-ca94-43c5-8165-87e26d08a4be","added_by":"auto","created_at":"2026-01-08 17:32:14","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":47836,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/29ac1bceb69f0604c34b4783.png"},{"id":100356575,"identity":"a694ce16-ad57-4959-8000-897c426f7cca","added_by":"auto","created_at":"2026-01-16 07:15:45","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":10824,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/0ecad0f05e0b74cf6c4c1ea8.png"},{"id":99830987,"identity":"929839a4-e484-4402-b719-53845ce2b1b4","added_by":"auto","created_at":"2026-01-08 17:32:14","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":65452,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/a76663c6eced23909d9b2499.png"},{"id":100356848,"identity":"0f9af397-98f6-4a64-a82f-e2ee777f7f52","added_by":"auto","created_at":"2026-01-16 07:17:46","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":328462,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/e38543e02125b1d4b2cdedee.png"},{"id":100356958,"identity":"b408a43e-71e4-4574-baa7-ee14d156bc65","added_by":"auto","created_at":"2026-01-16 07:18:07","extension":"xml","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":145269,"visible":true,"origin":"","legend":"","description":"","filename":"9d42161d153840fdb6b038ce5981dea11structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/ba6cbaffeaa1f9e5b1c91031.xml"},{"id":100356247,"identity":"81ae424e-dfa5-466f-b9b7-3fa6b3bde24f","added_by":"auto","created_at":"2026-01-16 06:58:22","extension":"html","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":166657,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1/45faae19adb5d977489fb769.html"},{"id":100376720,"identity":"e2a57193-fac9-4502-9478-ccbbcecfb87e","added_by":"auto","created_at":"2026-01-16 08:45:32","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":886905,"visible":true,"origin":"","legend":"","description":"","filename":"dfmirnet.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7996606/v1_covered_ad7672a6-7b8e-4336-b453-5cee6c4b7dd2.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"DFMIR-Net: Dual-Frequency Mamba Network for Single-Image Deraining","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":"the-journal-of-supercomputing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [The Journal of Supercomputing](https://www.springer.com/journal/11227)","snPcode":"11227","submissionUrl":"https://submission.nature.com/new-submission/11227/3","title":"The Journal of Supercomputing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Single-image deraining, Mamba, State-space models, Frequency-domain learning, Image restoration","lastPublishedDoi":"10.21203/rs.3.rs-7996606/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7996606/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSingle-image deraining aims to recover a clean image from inputs degraded by rain streaks, yet remains challenging because the streaks are cross-scale, anisotropic, and spatially complex. Existing approaches share several limitations: prior-driven methods are brittle under complex rain; CNNs, though efficient, are confined to local modeling and lack long-range dependencies; Transformers capture global context but have O(N\u0026sup2;) time and memory complexity with global self-attention and may oversmooth details; state-space models (SSMs) offer near-linear complexity but often underemphasize high-frequency details; and frequency-only designs risk ringing artifacts and unstable spatial\u0026ndash;spectral fusion. To address these issues, we combine global dependency modeling, lightweight local texture enhancement, and an explicit frequency prior under near-linear complexity, balancing global consistency with fine-detail fidelity. Building on this idea, we propose DFMIR-Net: selective SSMs provide efficient long-range modeling to complement CNNs and Transformers; a lightweight local enhancement mechanism strengthens textures and edges, improving performance in dense rain; and a rain-aware frequency module at the bottleneck employs a phase-preserving spectral representation with adaptive fusion so that spatial features and high-frequency cues jointly constrain restoration, mitigating SSMs\u0026rsquo; high-frequency underemphasis and stabilizing spatial\u0026ndash;spectral fusion. Experiments on Rain100L/H and Test100/1200/2800 show that DFMIR-Net achieves competitive or improved PSNR and SSIM compared with CNN-, Transformer-, and SSM-based methods, yields clearer visual details, and maintains near-linear computation and memory scaling.\u003c/p\u003e","manuscriptTitle":"DFMIR-Net: Dual-Frequency Mamba Network for Single-Image Deraining","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-08 17:32:09","doi":"10.21203/rs.3.rs-7996606/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-01-31T21:00:28+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-31T03:29:04+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-13T21:44:19+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"205388514947016617135792098436071729429","date":"2026-01-12T07:37:41+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"87575284090363017575408840204217781322","date":"2026-01-09T08:54:02+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"320925416162972608433834659376635500036","date":"2026-01-09T01:57:45+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"11820988249382401330083980240540653466","date":"2026-01-09T00:45:46+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-01-07T00:14:02+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-06T23:49:52+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-03T18:36:02+00:00","index":"","fulltext":""},{"type":"submitted","content":"The Journal of Supercomputing","date":"2025-10-31T08:56:01+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"the-journal-of-supercomputing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [The Journal of Supercomputing](https://www.springer.com/journal/11227)","snPcode":"11227","submissionUrl":"https://submission.nature.com/new-submission/11227/3","title":"The Journal of Supercomputing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"cc21afb8-7192-4d9c-a263-7ddfb44758f0","owner":[],"postedDate":"January 8th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-02-19T15:55:19+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-08 17:32:09","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7996606","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7996606","identity":"rs-7996606","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}