JVLGS: Joint Vision–Language Gas Leak Segmentation

JVLGS: Joint Vision–Language Gas Leak Segmentation | 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 JVLGS: Joint Vision–Language Gas Leak Segmentation Xinlong Zhao, Qixiang Pang, Shan Du This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9331073/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 11 You are reading this latest preprint version Abstract Gas leaks pose severe risks to human health and industrial safety. However, accurate and timely monitoring of gas leaks remains a major challenge. Existing vision-based methods using infrared (IR) imagery are limited by the inherently blurry and non-rigid nature of leak plumes, which reduces detection reliability and precision. To overcome these limitations, this paper proposes a Joint Vision–Language Gas leak Segmentation (JVLGS) framework that integrates the complementary strengths of visual and textual modalities to enhance gas leak segmentation. Recognizing that gas leaks are sporadic and many video frames contain no leakage, JVLGS incorporates an adaptive postprocessing module to effectively suppress false positives caused by noise and non-target objects—a common limitation of existing approaches. Extensive experiments across diverse industrial scenarios demonstrate that JVLGS significantly outperforms state-of-the-art gas leak segmentation methods. Furthermore, it achieves consistently strong performance under both supervised and few-shot learning settings, whereas competing methods typically perform well in only one setting or underperform in both. We publish our code at: https://github.com/GeekEagle/JVLGS. Gas leak segmentation Industrial safety monitoring Infrared video surveillance system Spatiotemporal perception Text-guided feature enhancement Vision language models (VLMs) Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 27 Apr, 2026 Reviews received at journal 21 Apr, 2026 Reviews received at journal 21 Apr, 2026 Reviews received at journal 14 Apr, 2026 Reviewers agreed at journal 14 Apr, 2026 Reviewers agreed at journal 13 Apr, 2026 Reviewers agreed at journal 13 Apr, 2026 Reviewers invited by journal 13 Apr, 2026 Editor assigned by journal 07 Apr, 2026 Submission checks completed at journal 07 Apr, 2026 First submitted to journal 06 Apr, 2026 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-9331073","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":624321809,"identity":"0038456b-863c-4516-9bcf-ee3ab0ca2f86","order_by":0,"name":"Xinlong Zhao","email":"","orcid":"","institution":"University of British Columbia","correspondingAuthor":false,"prefix":"","firstName":"Xinlong","middleName":"","lastName":"Zhao","suffix":""},{"id":624321810,"identity":"5810d33e-3122-45fa-b5ce-f398388715a0","order_by":1,"name":"Qixiang Pang","email":"","orcid":"","institution":"University of Central Missouri","correspondingAuthor":false,"prefix":"","firstName":"Qixiang","middleName":"","lastName":"Pang","suffix":""},{"id":624321811,"identity":"08e75733-62de-4132-8185-ada22609a93b","order_by":2,"name":"Shan Du","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAApUlEQVRIiWNgGAWjYDACCRDBxiAHYZCixZh0LYkNRGuRn9387AFD2eH0DbebHzD8qCFCi8GdY+YGDOcO5264c8yAsecYMVokEswkGNuAWm7kMDADXUiEw2akfwNpSTcAa/lHhBaGGzlgWxLAWhjbiHHYjZwyiYRz6YYzgX452NtHnMO2SXwos5bnu9388MGPb8Q4DAQSoPQBYjWMglEwCkbBKCAAAJdiNRKXMdZfAAAAAElFTkSuQmCC","orcid":"","institution":"University of British Columbia","correspondingAuthor":true,"prefix":"","firstName":"Shan","middleName":"","lastName":"Du","suffix":""}],"badges":[],"createdAt":"2026-04-06 07:38:34","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9331073/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9331073/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107398941,"identity":"1d9be0a5-63a9-4460-b1cc-757335960cd8","added_by":"auto","created_at":"2026-04-21 07:12:14","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":849053,"visible":true,"origin":"","legend":"","description":"","filename":"XinlongJVLGSVisualComputer.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9331073/v1_covered_9cf9ce9d-85ba-41b0-93e3-c256edcc1484.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"JVLGS: Joint Vision–Language Gas Leak Segmentation","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"the-visual-computer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"tvcj","sideBox":"Learn more about [The Visual Computer](http://link.springer.com/journal/371)","snPcode":"371","submissionUrl":"https://submission.nature.com/new-submission/371/3","title":"The Visual Computer","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Gas leak segmentation, Industrial safety monitoring, Infrared video surveillance system, Spatiotemporal perception, Text-guided feature enhancement, Vision language models (VLMs)","lastPublishedDoi":"10.21203/rs.3.rs-9331073/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9331073/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eGas leaks pose severe risks to human health and industrial safety. However, accurate and timely monitoring of gas leaks remains a major challenge. Existing vision-based methods using infrared (IR) imagery are limited by the inherently blurry and non-rigid nature of leak plumes, which reduces detection reliability and precision. To overcome these limitations, this paper proposes a Joint Vision–Language Gas leak Segmentation (JVLGS) framework that integrates the complementary strengths of visual and textual modalities to enhance gas leak segmentation. Recognizing that gas leaks are sporadic and many video frames contain no leakage, JVLGS incorporates an adaptive postprocessing module to effectively suppress false positives caused by noise and non-target objects—a common limitation of existing approaches. Extensive experiments across diverse industrial scenarios demonstrate that JVLGS significantly outperforms state-of-the-art gas leak segmentation methods. Furthermore, it achieves consistently strong performance under both supervised and few-shot learning settings, whereas competing methods typically perform well in only one setting or underperform in both. We publish our code at: https://github.com/GeekEagle/JVLGS.\u003c/p\u003e","manuscriptTitle":"JVLGS: Joint Vision–Language Gas Leak Segmentation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-21 07:09:55","doi":"10.21203/rs.3.rs-9331073/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-27T21:21:01+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-21T13:33:13+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-21T12:28:15+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-14T09:10:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"104125859979120947589348134952429489049","date":"2026-04-14T08:54:43+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"140392326679040372587239321671681865037","date":"2026-04-14T03:41:31+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"66188692130192187129217063952532450907","date":"2026-04-14T02:40:39+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-14T02:37:56+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-07T12:55:00+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-07T10:13:39+00:00","index":"","fulltext":""},{"type":"submitted","content":"The Visual Computer","date":"2026-04-06T07:32:28+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"the-visual-computer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"tvcj","sideBox":"Learn more about [The Visual Computer](http://link.springer.com/journal/371)","snPcode":"371","submissionUrl":"https://submission.nature.com/new-submission/371/3","title":"The Visual Computer","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"ae2cdad9-4902-49c4-a398-9a7cbdca4cf3","owner":[],"postedDate":"April 21st, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-04-27T21:23:41+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-21 07:09:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9331073","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9331073","identity":"rs-9331073","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}