MFTFF : multi-frame 3D object detection with temporal feature fusion

MFTFF : multi-frame 3D object detection with temporal feature fusion | 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 MFTFF : multi-frame 3D object detection with temporal feature fusion Xin Meng, Yuan Zhou, Min Zhang, Cui Wang, Jiahang Lv, Jonghyuk kim, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5878897/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 14 You are reading this latest preprint version Abstract In autonomous driving, 3D object detection is crucial. LiDAR continuously generate real-time point cloud, which serves as the foundation for 3D object detection. Traditional single-frame 3D detection methods typically rely on point cloud from individual time intervals. However, these methods often fail to fully utilize the temporal information in consecutive point cloud sequences, limiting their detection performance. In this work, the study introduces a novel architecture called the Multi-frame Temporal Feature Fusion Network (MFTFF). MFTFF enhances the features at the current time by effectively integrating the features from the point cloud sequence. The network includes the Multi-frame Feature Conversion and Alignment Module (MFCA). This module aligns different point cloud frames in the time series to a unified perspective. In addition, we introduce an efficient Multi-frame Feature Fusion Module(MFFF). This module performs a coarse-grained fusion through concatenation and a fine-grained fusion by using a temporal attention mechanism. This fusion strategy preserves the essential temporal information and improves the 3D object detection performance effectively. After the fusion, the final features are stored in the memory bank for continuous updates and iterations. This mechanism provides rich feature information for subsequent detection tasks. On the nuScenes dataset, MFTFF achieved 68.6% in NDS and 62.2% in mAP. Compared to the baseline method, MFTFF improved NDS by 1.3% and mAP by 1.9%. 3D object detection lidar point cloud temporal fusion multi-frame Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 08 Apr, 2025 Reviews received at journal 17 Mar, 2025 Reviews received at journal 14 Mar, 2025 Reviewers agreed at journal 11 Mar, 2025 Reviewers agreed at journal 08 Mar, 2025 Reviewers agreed at journal 07 Mar, 2025 Reviewers agreed at journal 07 Mar, 2025 Reviewers agreed at journal 06 Mar, 2025 Reviewers agreed at journal 06 Mar, 2025 Reviewers agreed at journal 05 Mar, 2025 Reviewers invited by journal 05 Mar, 2025 Editor assigned by journal 24 Jan, 2025 Submission checks completed at journal 24 Jan, 2025 First submitted to journal 22 Jan, 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-5878897","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":406910296,"identity":"ab9432f9-db1e-4da2-add7-38dbd89ad794","order_by":0,"name":"Xin Meng","email":"","orcid":"","institution":"Changchun University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Meng","suffix":""},{"id":406910297,"identity":"2ffc9f56-362e-424d-ab2a-a0b5788cdf16","order_by":1,"name":"Yuan Zhou","email":"","orcid":"","institution":"Leapmotor","correspondingAuthor":false,"prefix":"","firstName":"Yuan","middleName":"","lastName":"Zhou","suffix":""},{"id":406910298,"identity":"7d37a109-5792-425c-ae63-d08ee14dffa1","order_by":2,"name":"Min Zhang","email":"","orcid":"","institution":"Changchun University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Min","middleName":"","lastName":"Zhang","suffix":""},{"id":406910299,"identity":"627bd0a0-3e45-4385-845c-1e42e10e73bb","order_by":3,"name":"Cui Wang","email":"","orcid":"","institution":"Zhongshan Institute of Changchun University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Cui","middleName":"","lastName":"Wang","suffix":""},{"id":406910300,"identity":"2b8789f6-ad39-4cd8-9ad8-b62cb0c3a863","order_by":4,"name":"Jiahang Lv","email":"","orcid":"","institution":"Changchun University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Jiahang","middleName":"","lastName":"Lv","suffix":""},{"id":406910301,"identity":"20fec472-213f-4f6a-87e9-1b57d3ba8a53","order_by":5,"name":"Jonghyuk kim","email":"","orcid":"","institution":"Naif Arab University for Security Sciences, Kingdom of Saudi Arabia","correspondingAuthor":false,"prefix":"","firstName":"Jonghyuk","middleName":"","lastName":"kim","suffix":""},{"id":406910302,"identity":"a5f37d08-585a-4269-a162-8997245a36f5","order_by":6,"name":"Shifeng Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2klEQVRIiWNgGAWjYDAC5gNgSg7GJUILWwKYMgZixgaStCQ2EK1Fvo2BTeLnjtr0/hnpzx8wVFgnNrCfPYBXi8ExBjbJ3jPHc2fcyDFsYDiTntjAk5eAX4t8A9sN3rZjuRskchgbGNsOJzZI8BgQdNjNv23H0g0k0h82MP4jQgsD0GG3edtqEgwkEgwbGBuI0GJwjLH9t2zbAcMZZ94Yzkg4lm7cxpNDyGHMhw3fttXJ87enP/jwocZatp/9DAGHgWOD4TCEnQDEbATUw0AdkepGwSgYBaNgRAIA7qxElEPEwJAAAAAASUVORK5CYII=","orcid":"","institution":"Changchun University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Shifeng","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2025-01-22 08:23:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5878897/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5878897/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":74956234,"identity":"eadb3223-3f85-4390-a7e8-df69733e1c4e","added_by":"auto","created_at":"2025-01-28 17:37:11","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":846396,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5878897/v1_covered_b9026268-729f-4e52-a7b0-bd0d0eb2bac7.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"MFTFF : multi-frame 3D object detection with temporal feature fusion","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":"3D object detection, lidar point cloud, temporal fusion, multi-frame","lastPublishedDoi":"10.21203/rs.3.rs-5878897/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5878897/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn autonomous driving, 3D object detection is crucial. LiDAR continuously generate real-time point cloud, which serves as the foundation for 3D object detection. Traditional single-frame 3D detection methods typically rely on point cloud from individual time intervals. However, these methods often fail to fully utilize the temporal information in consecutive point cloud sequences, limiting their detection performance. In this work, the study introduces a novel architecture called the Multi-frame Temporal Feature Fusion Network (MFTFF). MFTFF enhances the features at the current time by effectively integrating the features from the point cloud sequence. The network includes the Multi-frame Feature Conversion and Alignment Module (MFCA). This module aligns different point cloud frames in the time series to a unified perspective. In addition, we introduce an efficient Multi-frame Feature Fusion Module(MFFF). This module performs a coarse-grained fusion through concatenation and a fine-grained fusion by using a temporal attention mechanism. This fusion strategy preserves the essential temporal information and improves the 3D object detection performance effectively. After the fusion, the final features are stored in the memory bank for continuous updates and iterations. This mechanism provides rich feature information for subsequent detection tasks. On the nuScenes dataset, MFTFF achieved 68.6% in NDS and 62.2% in mAP. Compared to the baseline method, MFTFF improved NDS by 1.3% and mAP by 1.9%.\u003c/p\u003e","manuscriptTitle":"MFTFF : multi-frame 3D object detection with temporal feature fusion","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-28 17:21:06","doi":"10.21203/rs.3.rs-5878897/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-04-08T20:54:12+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-03-17T05:22:42+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-03-14T22:32:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"202009709233073950036639701363167545151","date":"2025-03-11T21:20:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"60412865358858856117552494798004623430","date":"2025-03-08T06:02:02+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"209678226890050453001188194643815340495","date":"2025-03-08T03:32:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"85886173488100515748477863537785640679","date":"2025-03-07T10:48:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"30712240411623917055870296422328768835","date":"2025-03-06T13:19:55+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"64246598930464706326935468780604928326","date":"2025-03-06T11:32:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"164927084195943758765245002418912646730","date":"2025-03-06T04:05:14+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-03-06T02:53:50+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-01-25T03:28:54+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-01-25T03:28:39+00:00","index":"","fulltext":""},{"type":"submitted","content":"The Journal of Supercomputing","date":"2025-01-22T08:14:45+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":"2309bfcc-ca6b-47ef-badc-df3272e5d5ab","owner":[],"postedDate":"January 28th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2025-05-16T23:08:12+00:00","versionOfRecord":[],"versionCreatedAt":"2025-01-28 17:21:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5878897","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5878897","identity":"rs-5878897","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}