UniTracker:Transformer Based CrossUnihead for Multi-object Tracking

UniTracker:Transformer Based CrossUnihead for Multi-object Tracking | 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 UniTracker:Transformer Based CrossUnihead for Multi-object Tracking Fan Wu, Yifeng Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4002528/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 15 Jul, 2024 Read the published version in Journal of Real-Time Image Processing → Version 1 posted 10 You are reading this latest preprint version Abstract In recent years, tracking-by-detection (TBD) has emerged as the predominant approach for Multi-object Tracking (MOT). Most TBD algorithms typically employ separate branch heads to handle the coarse feature representations extracted from the backbone network. However, this approach often leads to suboptimal model performance. This is due to the presence of feature conflicts among multiple tasks on one hand, and the lack of information interaction among these tasks on the other. Inspired by CSTrack and Unihead, we have combined the feature decoupling module REN with the transformer-based task interaction module CIT to form CrossUnihead which is a plug and play module. This module not only effectively achieves task-based feature decoupling, but also facilitates information interaction among different tasks. The MOT algorithm based on CrossUnihead, named UniTracker, demonstrates impressive performance on the MOT dataset when compared to other advanced methods and is capable of achieving real-time tracking. Feature Decoupling Task Interaction Module CrossUnihead Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 15 Jul, 2024 Read the published version in Journal of Real-Time Image Processing → Version 1 posted Editorial decision: Revision requested 21 May, 2024 Reviews received at journal 19 May, 2024 Reviewers agreed at journal 29 Apr, 2024 Reviews received at journal 20 Mar, 2024 Reviewers agreed at journal 19 Mar, 2024 Reviewers agreed at journal 06 Mar, 2024 Reviewers invited by journal 04 Mar, 2024 Submission checks completed at journal 02 Mar, 2024 Editor assigned by journal 02 Mar, 2024 First submitted to journal 01 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-4002528","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":276425862,"identity":"983f6c65-9259-43ee-aa01-a41b6d0e3a90","order_by":0,"name":"Fan Wu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8klEQVRIie3RsWrDMBCA4RMCZTmS1aLkHQ4E6dCQDn0RiYC7+AEyKhTcxZSuhT6EH0GgwYsfwCFDbQLJkqFjS4eWeAxFTrcO+sfjvuUOIBb7h01G63Wrab6YjGw/YHaIyMJ7alfpUhbuQkJNmsq29qxs9IUEXEaJyT2nzaHbIcynpeP7NiSYrYlMfj+W20wphFSVTlxTiHBWkDb5DR9vM3GF4E3pUCQhIjiSMzlndlMdvhC+hwkKYayu79hzAzOO4IZJgtyD7o+cKflKS/XixSxIbt+6x8+P/pVV935cLaZP1cM+SM46nYr/YT8Wi8Viv/cDuShM5fiLgDUAAAAASUVORK5CYII=","orcid":"","institution":"Southeast University","correspondingAuthor":true,"prefix":"","firstName":"Fan","middleName":"","lastName":"Wu","suffix":""},{"id":276425863,"identity":"16eb88f3-f769-4369-b884-d095b7e8ad16","order_by":1,"name":"Yifeng Zhang","email":"","orcid":"","institution":"Southeast University","correspondingAuthor":false,"prefix":"","firstName":"Yifeng","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2024-03-01 08:04:53","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4002528/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4002528/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s11554-024-01514-9","type":"published","date":"2024-07-15T16:05:18+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":61596469,"identity":"661d6531-db79-4d35-b7af-8053e0398b79","added_by":"auto","created_at":"2024-08-01 17:27:49","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":415918,"visible":true,"origin":"","legend":"","description":"","filename":"FanWuUniTracker18851892889220210939seu.edu.cn.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4002528/v1_covered_c7430218-6e7d-4f93-a252-eca1816f9c73.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"UniTracker:Transformer Based CrossUnihead for Multi-object Tracking","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":"journal-of-real-time-image-processing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"rtip","sideBox":"Learn more about [Journal of Real-Time Image Processing](http://link.springer.com/journal/11554)","snPcode":"11554","submissionUrl":"https://submission.nature.com/new-submission/11554/3","title":"Journal of Real-Time Image Processing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Feature Decoupling, Task Interaction Module, CrossUnihead","lastPublishedDoi":"10.21203/rs.3.rs-4002528/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4002528/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":" In recent years, tracking-by-detection (TBD) has emerged as the predominant approach for \n Multi-object Tracking (MOT). Most TBD algorithms typically employ separate branch heads to handle\n the coarse feature representations extracted from the backbone network. However, this approach often leads to suboptimal model performance.\n This is due to the presence of feature conflicts among multiple tasks on one hand, and the lack of information interaction among these tasks on the other. \n Inspired by CSTrack and Unihead, we have combined the feature decoupling module REN with the transformer-based task interaction module CIT to form CrossUnihead which is a plug and play module. \n This module not only effectively achieves task-based feature decoupling, but also facilitates information interaction among different tasks. \n The MOT algorithm based on CrossUnihead, named UniTracker, demonstrates impressive performance on the MOT dataset when compared to other advanced methods and is capable of achieving real-time tracking.","manuscriptTitle":"UniTracker:Transformer Based CrossUnihead for Multi-object Tracking","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-03-06 06:45:52","doi":"10.21203/rs.3.rs-4002528/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-05-21T11:55:28+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-20T02:12:10+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"162952009440615821347959552268210154809","date":"2024-04-29T12:17:35+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-03-20T04:02:18+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"1960eeb3-d4be-4aba-9ea9-356589505b38","date":"2024-03-19T20:15:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"11bc8081-21be-426b-bbea-d55b0a19577d","date":"2024-03-06T10:50:31+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-03-04T10:46:25+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-03-02T09:54:23+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-03-02T09:54:23+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Real-Time Image Processing","date":"2024-03-01T07:56:25+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-real-time-image-processing","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"rtip","sideBox":"Learn more about [Journal of Real-Time Image Processing](http://link.springer.com/journal/11554)","snPcode":"11554","submissionUrl":"https://submission.nature.com/new-submission/11554/3","title":"Journal of Real-Time Image Processing","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"648c402f-88a3-4d49-b4ad-24ea54ebfe82","owner":[],"postedDate":"March 6th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-08-01T17:12:05+00:00","versionOfRecord":{"articleIdentity":"rs-4002528","link":"https://doi.org/10.1007/s11554-024-01514-9","journal":{"identity":"journal-of-real-time-image-processing","isVorOnly":false,"title":"Journal of Real-Time Image Processing"},"publishedOn":"2024-07-15 16:05:18","publishedOnDateReadable":"July 15th, 2024"},"versionCreatedAt":"2024-03-06 06:45:52","video":"","vorDoi":"10.1007/s11554-024-01514-9","vorDoiUrl":"https://doi.org/10.1007/s11554-024-01514-9","workflowStages":[]},"version":"v1","identity":"rs-4002528","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4002528","identity":"rs-4002528","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}