SCM: Semantic Segmentation with Dual-Stream Semantic Synergy under Adverse Weather Conditions | 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 SCM: Semantic Segmentation with Dual-Stream Semantic Synergy under Adverse Weather Conditions Shuochen Tian, Jian Pang, Jin Wang, Bingfeng Zhang, Weifeng Liu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7447174/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 04 Dec, 2025 Read the published version in Multimedia Systems → Version 1 posted 9 You are reading this latest preprint version Abstract Unsupervised domain adaptation (UDA) aims to adapt a model trained on a source domain (\eg, clear weather data), to a target domain (\eg, adverse weather data), without the need for annotations in the target domain. Recent semantic segmentation models perform well under clear weather conditions but struggle with adverse weather conditions, which leading to poor segmentation results. There are two routes for these methods: designing a network for feature mapping or extracting domain-invariant features by additional modal information, but it is difficult to learn robust domain-invariant features. To address these issues, we introduce CLIP and design a dual-stream feature fusion network (DSFF), and Our DSFF allows CLIP and CNN encoders to benefit from each other, which increases model performance. This dual-stream design enables mutual enhancement between the two encoders, the frozen CLIP supplies the CNN encoder with high-level semantic cues relevant to the target domain, while the CNN encoder helps CLIP to reduce weather noise. Extensive experiments show that our method significantly outperforms other methods, achieving a mIoU of 59.0 on Cityscapes to ACDC. Unsupervised Domain Adaptation Semantic Segmentation Dual-stream Feature Fusion Module domain-invariant features CLIP Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 04 Dec, 2025 Read the published version in Multimedia Systems → Version 1 posted Editorial decision: Revision requested 26 Sep, 2025 Reviews received at journal 26 Sep, 2025 Reviewers agreed at journal 26 Sep, 2025 Reviews received at journal 31 Aug, 2025 Reviewers agreed at journal 28 Aug, 2025 Reviewers invited by journal 26 Aug, 2025 Editor assigned by journal 26 Aug, 2025 Submission checks completed at journal 25 Aug, 2025 First submitted to journal 24 Aug, 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-7447174","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":508217457,"identity":"6494a242-4705-4ced-96a6-bfb191b94a60","order_by":0,"name":"Shuochen Tian","email":"","orcid":"","institution":"China University of Petroleum (East China)","correspondingAuthor":false,"prefix":"","firstName":"Shuochen","middleName":"","lastName":"Tian","suffix":""},{"id":508217458,"identity":"3deb5333-3415-4c06-9507-8be6d04d4e22","order_by":1,"name":"Jian Pang","email":"","orcid":"","institution":"China University of Petroleum (East China)","correspondingAuthor":false,"prefix":"","firstName":"Jian","middleName":"","lastName":"Pang","suffix":""},{"id":508217459,"identity":"fa0e64fb-00c3-47c6-b684-df981487baaa","order_by":2,"name":"Jin Wang","email":"","orcid":"","institution":"China University of Petroleum (East China)","correspondingAuthor":false,"prefix":"","firstName":"Jin","middleName":"","lastName":"Wang","suffix":""},{"id":508217460,"identity":"38a796bf-04b2-4855-80ed-9eeefeda5081","order_by":3,"name":"Bingfeng Zhang","email":"","orcid":"","institution":"China University of Petroleum (East China)","correspondingAuthor":false,"prefix":"","firstName":"Bingfeng","middleName":"","lastName":"Zhang","suffix":""},{"id":508217461,"identity":"e54fe3f3-62bf-45a2-9724-98759f9f793f","order_by":4,"name":"Weifeng Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAq0lEQVRIiWNgGAWjYBACPmYgUVEB4UgQpYUNpOXMGZK0gIizbSRpYecxkzg473AefwPzwds8DHZ5RDgMpGXb4WKJA2zJ1jwMycVEabn9cdvhxA0MPGbSPAwHEhuI0XLj4ByQFv5vpGhpANvCRqwWtvIfB46lF0scZjO2nGOQTFgLP//hzQYHaqzz+NubH954U2FHWAsMJDCA4pTBgFj1YC2jYBSMglEwCnABAIZdNJnxxQkFAAAAAElFTkSuQmCC","orcid":"","institution":"China University of Petroleum (East China)","correspondingAuthor":true,"prefix":"","firstName":"Weifeng","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2025-08-24 15:38:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7447174/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7447174/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00530-025-02092-x","type":"published","date":"2025-12-04T15:57:52+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":97723879,"identity":"d71bc34a-ad83-4090-8335-a87c6a29cf54","added_by":"auto","created_at":"2025-12-08 16:09:02","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":719721,"visible":true,"origin":"","legend":"","description":"","filename":"ChinammID1351.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7447174/v1_covered_cf879d05-aead-4ecc-97c5-49562250e067.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"SCM: Semantic Segmentation with Dual-Stream Semantic Synergy under Adverse Weather Conditions","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":"multimedia-systems","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mmsj","sideBox":"Learn more about [Multimedia Systems](http://link.springer.com/journal/530)","snPcode":"530","submissionUrl":"https://submission.nature.com/new-submission/530/3","title":"Multimedia Systems","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Unsupervised Domain Adaptation Semantic Segmentation, Dual-stream Feature Fusion Module, domain-invariant features, CLIP","lastPublishedDoi":"10.21203/rs.3.rs-7447174/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7447174/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eUnsupervised domain adaptation (UDA) aims to adapt a model trained on a source domain (\\eg, clear weather data), to a target domain (\\eg, adverse weather data), without the need for annotations in the target domain. Recent semantic segmentation models perform well under clear weather conditions but struggle with adverse weather conditions, which leading to poor segmentation results. There are two routes for these methods: designing a network for feature mapping or extracting domain-invariant features by additional modal information, but it is difficult to learn robust domain-invariant features. To address these issues, we introduce CLIP and design a dual-stream feature fusion network (DSFF), and Our DSFF allows CLIP and CNN encoders to benefit from each other, which increases model performance. This dual-stream design enables mutual enhancement between the two encoders, the frozen CLIP supplies the CNN encoder with high-level semantic cues relevant to the target domain, while the CNN encoder helps CLIP to reduce weather noise. Extensive experiments show that our method significantly outperforms other methods, achieving a mIoU of 59.0 on Cityscapes to ACDC.\u003c/p\u003e","manuscriptTitle":"SCM: Semantic Segmentation with Dual-Stream Semantic Synergy under Adverse Weather Conditions","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-03 07:56:00","doi":"10.21203/rs.3.rs-7447174/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-27T02:54:27+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-26T14:51:26+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"235861674037773211039418470905193037999","date":"2025-09-26T07:31:18+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-31T08:16:23+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"108349836684400870756585683715655394626","date":"2025-08-29T03:02:53+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-27T02:30:54+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-26T07:57:51+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-25T14:20:54+00:00","index":"","fulltext":""},{"type":"submitted","content":"Multimedia Systems","date":"2025-08-24T15:33:11+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"
[email protected]","identity":"multimedia-systems","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mmsj","sideBox":"Learn more about [Multimedia Systems](http://link.springer.com/journal/530)","snPcode":"530","submissionUrl":"https://submission.nature.com/new-submission/530/3","title":"Multimedia Systems","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"123daddd-65f1-4084-b9e9-45aa04e20679","owner":[],"postedDate":"September 3rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-08T16:02:02+00:00","versionOfRecord":{"articleIdentity":"rs-7447174","link":"https://doi.org/10.1007/s00530-025-02092-x","journal":{"identity":"multimedia-systems","isVorOnly":false,"title":"Multimedia Systems"},"publishedOn":"2025-12-04 15:57:52","publishedOnDateReadable":"December 4th, 2025"},"versionCreatedAt":"2025-09-03 07:56:00","video":"","vorDoi":"10.1007/s00530-025-02092-x","vorDoiUrl":"https://doi.org/10.1007/s00530-025-02092-x","workflowStages":[]},"version":"v1","identity":"rs-7447174","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7447174","identity":"rs-7447174","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.