GCL-BEV: Enhancing Pure Vision 3D Detection with Motion Priors and View-Consistency Learning

GCL-BEV: Enhancing Pure Vision 3D Detection with Motion Priors and View-Consistency Learning | 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 GCL-BEV: Enhancing Pure Vision 3D Detection with Motion Priors and View-Consistency Learning Zhipeng Qi, Zhijun Xie, Jing Xu, Rui Wang, Ming Jin This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9196712/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 6 You are reading this latest preprint version Abstract Temporal fusion has become a de facto standard in vision-centric Bird’s-Eye-View (BEV) perception, enabling velocity estimation and occlusion mitigation. However, existing paradigms typically rely on rigid geometric alignment (e.g., ego-pose warping) to aggregate historical features. We identify that this assumption is fragile: under aggressive ego-motion, such as rapid turning, the non-linear distortion of visual features leads to significant spatial misalignment, causing prediction jitter and feature smearing. To bridge this gap, we propose GCL-BEV, a robust detection framework that enforces geometric consistency through both architectural evolution and optimization constraints. First, we introduce a Geometric-Aware Feature Enhancement (GAFE) module. Unlike standard deformable convolutions that infer offsets from visual appearance, GAFE explicitly utilizes kinematic priors (ego-motion) to guide the dynamic deformation of the receptive field, ensuring feature alignment before temporal fusion. Second, we propose a View-Consistency Learning (VCL) objective. Formulated as a Siamese equivariance constraint, VCL compels the backbone to learn rotation-invariant representations during training, enhancing robustness against viewpoint perturbations with strictly zero inference overhead. Extensive experiments on the nuScenes dataset demonstrate that GCL-BEV achieves state-of-the-art performance among ResNet-101 based methods (57.8% NDS, 46.2% mAP). Crucially, our method significantly reduces Orientation Error (mAOE) by 5.4% compared to the baseline, validating its superiority in maintaining geometric stability under complex driving maneuvers. Autonomous Driving BEV Perception Temporal Fusion Geometric Equivariance Deformable Convolution. Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 10 Apr, 2026 Reviewers agreed at journal 02 Apr, 2026 Reviewers invited by journal 27 Mar, 2026 Editor assigned by journal 24 Mar, 2026 Submission checks completed at journal 24 Mar, 2026 First submitted to journal 23 Mar, 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-9196712","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":613380647,"identity":"43cd75bb-9b41-4169-8633-c93bfe2a3606","order_by":0,"name":"Zhipeng Qi","email":"","orcid":"","institution":"Ningbo University","correspondingAuthor":false,"prefix":"","firstName":"Zhipeng","middleName":"","lastName":"Qi","suffix":""},{"id":613380650,"identity":"c73fac08-abe9-4b71-be66-df2454f4af7c","order_by":1,"name":"Zhijun Xie","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBUlEQVRIie3PoWoDMRjA8a8EMpPb2TvRvsEgR6Cq9FkaAlUTHTMVFQeBnTmoPTH2DFE3ppYSyMygdnJlLxComZhYjqvrrj05aP5EhfySfACh0H8MNYtigFjqzC39zhXAIO9FEju7T94BCDpHWtV0S1300IPQN2T2i8U1fwadqfRpwl8R2uwqmI5uOlwq8YxVFPMXmQt6V895ibDIFAg21n+TGBEqiCdKa0vT2nhCxuknaF53EIxiZ1rCi+/osQfxrwxkSwTQKD8QdYL4WZj/G2bqwwJL7JyVxs9S0e5Z6NZ87cmPHart2mVuNRkWhdzsyuV01EUO2aOrTh5vWp09EQqFQhfcL8IcWlSw53BFAAAAAElFTkSuQmCC","orcid":"","institution":"Ningbo University","correspondingAuthor":true,"prefix":"","firstName":"Zhijun","middleName":"","lastName":"Xie","suffix":""},{"id":613380655,"identity":"35fbd8b1-0056-43dc-ad25-8ede36320cc1","order_by":2,"name":"Jing Xu","email":"","orcid":"","institution":"Ningbo BIRD E-LINK Technology Co., Ltd.","correspondingAuthor":false,"prefix":"","firstName":"Jing","middleName":"","lastName":"Xu","suffix":""},{"id":613380657,"identity":"a71fb257-5b42-45cd-b7fa-6dda27590d81","order_by":3,"name":"Rui Wang","email":"","orcid":"","institution":"Ningbo University","correspondingAuthor":false,"prefix":"","firstName":"Rui","middleName":"","lastName":"Wang","suffix":""},{"id":613380659,"identity":"76460078-b54e-4e6c-94e9-95d003a7c94e","order_by":4,"name":"Ming Jin","email":"","orcid":"","institution":"Ningbo University","correspondingAuthor":false,"prefix":"","firstName":"Ming","middleName":"","lastName":"Jin","suffix":""}],"badges":[],"createdAt":"2026-03-23 07:09:36","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9196712/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9196712/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105904182,"identity":"90be958c-a4df-427f-9c28-e9f26ae2ba7c","added_by":"auto","created_at":"2026-04-01 10:05:53","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1279959,"visible":true,"origin":"","legend":"","description":"","filename":"GCLBEV.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9196712/v1_covered_b8bd944f-784a-4bf3-9621-83e9f6e9c70f.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"GCL-BEV: Enhancing Pure Vision 3D Detection with Motion Priors and View-Consistency Learning","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":"journal-of-king-saud-university-computer-and-information-sciences","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Journal of King Saud University Computer and Information Sciences](https://link.springer.com/journal/44443)","snPcode":"44443","submissionUrl":"https://submission.springernature.com/new-submission/44443/3","title":"Journal of King Saud University Computer and Information Sciences","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Open","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Autonomous Driving, BEV Perception, Temporal Fusion, Geometric Equivariance, Deformable Convolution.","lastPublishedDoi":"10.21203/rs.3.rs-9196712/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9196712/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTemporal fusion has become a de facto standard in vision-centric Bird\u0026rsquo;s-Eye-View (BEV) perception, enabling velocity estimation and occlusion mitigation. However, existing paradigms typically rely on rigid geometric alignment (e.g., ego-pose warping) to aggregate historical features. We identify that this assumption is fragile: under aggressive ego-motion, such as rapid turning, the non-linear distortion of visual features leads to significant spatial misalignment, causing prediction jitter and feature smearing. To bridge this gap, we propose GCL-BEV, a robust detection framework that enforces geometric consistency through both architectural evolution and optimization constraints. First, we introduce a Geometric-Aware Feature Enhancement (GAFE) module. Unlike standard deformable convolutions that infer offsets from visual appearance, GAFE explicitly utilizes kinematic priors (ego-motion) to guide the dynamic deformation of the receptive field, ensuring feature alignment before temporal fusion. Second, we propose a View-Consistency Learning (VCL) objective. Formulated as a Siamese equivariance constraint, VCL compels the backbone to learn rotation-invariant representations during training, enhancing robustness against viewpoint perturbations with strictly zero inference overhead. Extensive experiments on the nuScenes dataset demonstrate that GCL-BEV achieves state-of-the-art performance among ResNet-101 based methods (57.8% NDS, 46.2% mAP). Crucially, our method significantly reduces Orientation Error (mAOE) by 5.4% compared to the baseline, validating its superiority in maintaining geometric stability under complex driving maneuvers.\u003c/p\u003e","manuscriptTitle":"GCL-BEV: Enhancing Pure Vision 3D Detection with Motion Priors and View-Consistency Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-31 11:06:41","doi":"10.21203/rs.3.rs-9196712/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-04-10T17:41:43+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"129216048815762463259954775216191410754","date":"2026-04-02T15:54:59+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-27T14:24:11+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-24T04:23:17+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-24T04:22:55+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of King Saud University Computer and Information Sciences","date":"2026-03-23T06:52:52+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-king-saud-university-computer-and-information-sciences","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Journal of King Saud University Computer and Information Sciences](https://link.springer.com/journal/44443)","snPcode":"44443","submissionUrl":"https://submission.springernature.com/new-submission/44443/3","title":"Journal of King Saud University Computer and Information Sciences","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Open","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"58841bbc-e40b-484e-b1b9-82afbfae0f07","owner":[],"postedDate":"March 31st, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-03-31T11:06:41+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-31 11:06:41","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9196712","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9196712","identity":"rs-9196712","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}