CenterPoint-SE: A Single-Stage Anchor-Free 3-D Object Detection Algorithm With Spatial Awareness Enhancement

Real-time and accurate 3-D object detection is one of the foundational technologies for environmental perception in autonomous vehicles. However, the existing second-stage anchor-based 3-D object detection algorithms have high accuracy, but they are challenging in terms of computation complexity and...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	IEEE transactions on intelligent transportation systems Jg. 24; H. 10; S. 10760 - 10773
Hauptverfasser:	Wang, Hai, Tao, Le, Cai, Yingfeng, Chen, Long, Li, Yicheng, Sotelo, Miguel Angel, Li, Zhixiong
Format:	Journal Article
Sprache:	Englisch
Veröffentlicht:	New York IEEE 01.10.2023 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:	3D object detection Accuracy Algorithms anchor-free Autonomous vehicles Computer networks Detection algorithms Feature extraction Location awareness Object detection Object recognition Perception point cloud Point cloud compression single-stage space feature enhancement Training
ISSN:	1524-9050, 1558-0016
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Abstract	Real-time and accurate 3-D object detection is one of the foundational technologies for environmental perception in autonomous vehicles. However, the existing second-stage anchor-based 3-D object detection algorithms have high accuracy, but they are challenging in terms of computation complexity and latency. Due to poor perception of spatial features, the accuracy of the existing single-stage anchor-free detection algorithms with low latency are difficult to be implemented into autonomous vehicles. Therefore, we focus on enhancing the spatial perception ability of the anchor-free detection network based on CenterPoints. In this paper, we propose a single-stage anchor-free 3-D object detector CenterPoint-Space-Enhancement (CenterPoint-SE) algorithm and construct an efficient 3-D backbone network to extract fine-grained spatial geometric features by introducing a spatial attention mechanism and residual structure. At the same time, a powerful spatial semantic feature fusion module, the enhancement of feature fusion (EF-Fusion), is designed. In addition, we add a lightweight IoU prediction branch to improve the algorithm's perception of various object sizes. Finally, we add a foreground point segmentation auxiliary training branch to enable the 3-D backbone to obtain object boundary features. We use the ONCE dataset to train and validate the proposed model, and the results showed that the proposed CenterPoint-SE achieves 70.33 mAP and an inference speed of 17.15 FPS, outperforming other methods.
AbstractList	Real-time and accurate 3-D object detection is one of the foundational technologies for environmental perception in autonomous vehicles. However, the existing second-stage anchor-based 3-D object detection algorithms have high accuracy, but they are challenging in terms of computation complexity and latency. Due to poor perception of spatial features, the accuracy of the existing single-stage anchor-free detection algorithms with low latency are difficult to be implemented into autonomous vehicles. Therefore, we focus on enhancing the spatial perception ability of the anchor-free detection network based on CenterPoints. In this paper, we propose a single-stage anchor-free 3-D object detector CenterPoint-Space-Enhancement (CenterPoint-SE) algorithm and construct an efficient 3-D backbone network to extract fine-grained spatial geometric features by introducing a spatial attention mechanism and residual structure. At the same time, a powerful spatial semantic feature fusion module, the enhancement of feature fusion (EF-Fusion), is designed. In addition, we add a lightweight IoU prediction branch to improve the algorithm’s perception of various object sizes. Finally, we add a foreground point segmentation auxiliary training branch to enable the 3-D backbone to obtain object boundary features. We use the ONCE dataset to train and validate the proposed model, and the results showed that the proposed CenterPoint-SE achieves 70.33 mAP and an inference speed of 17.15 FPS, outperforming other methods.
Author	Li, Yicheng Chen, Long Wang, Hai Sotelo, Miguel Angel Cai, Yingfeng Tao, Le Li, Zhixiong
Author_xml	– sequence: 1 givenname: Hai orcidid: 0000-0002-9136-8091 surname: Wang fullname: Wang, Hai email: wanghai1019@163.com organization: School of Automotive and Traffic Engineering, Jiangsu University, Zhenjiang, China – sequence: 2 givenname: Le surname: Tao fullname: Tao, Le email: 1747524097@qq.com organization: School of Automotive and Traffic Engineering, Jiangsu University, Zhenjiang, China – sequence: 3 givenname: Yingfeng orcidid: 0000-0002-0633-9887 surname: Cai fullname: Cai, Yingfeng email: caicaixiao0304@126.com organization: Automotive Engineering Research Institute, Jiangsu University, Zhenjiang, China – sequence: 4 givenname: Long orcidid: 0000-0002-2079-3867 surname: Chen fullname: Chen, Long email: chenlong@ujs.edu.cn organization: Automotive Engineering Research Institute, Jiangsu University, Zhenjiang, China – sequence: 5 givenname: Yicheng surname: Li fullname: Li, Yicheng email: liyucheng070@163.com organization: Automotive Engineering Research Institute, Jiangsu University, Zhenjiang, China – sequence: 6 givenname: Miguel Angel orcidid: 0000-0001-8809-2103 surname: Sotelo fullname: Sotelo, Miguel Angel email: miguel.sotelo@uah.es organization: Department of Computer Engineering, University of Alcal, Alcal de Henares, Madrid, Spain – sequence: 7 givenname: Zhixiong orcidid: 0000-0002-7265-0008 surname: Li fullname: Li, Zhixiong email: zhixiong.li@yonsei.ac.kr organization: Yonsei Frontier Laboratory, Yonsei University, Seoul, Republic of Korea
BookMark	eNp9UE1LxDAQDaLg5w8QPAQ8Z03SJm28lXX9AEGhKx7LtJ3dzdJN1zSL-O9NWQ_iQRhmhuG9eTPvlBy63iEhl4JPhODmZv40LyeSy2SSyCzJRXZAToRSOeNc6MOxlykzXPFjcjoM6zhNlRAnpJuiC-hfe-sCK2e3tKCldcsOWRlgibRwzar37N4j0oTd0Zd6jU2gdxhisb2jRbfsvQ2rDX2PmZZbCBY6WnyCR4fDQGduBa7BTdQ5J0cL6Aa8-Kln5O1-Np8-sueXh6dp8cwaadLAdJ7Wqm0XBgDyHIyuc6ilyWChNQreYg3QtqkWLWRKG9kkBngM1FpoTFVyRq73e7e-_9jhEKp1v_MuSlYyz2SaqEzlEZXtUY3vh8HjompsgPGp4MF2leDVaG01WluN1lY_1kam-MPcersB__Uv52rPsYj4Cy_iRUYl3-6Whp4
CODEN	ITISFG
CitedBy_id	crossref_primary_10_1109_TIM_2023_3338662 crossref_primary_10_1109_TITS_2024_3429139
Cites_doi	10.1109/CVPR.2018.00472 10.1109/TETCI.2023.3235381 10.1016/j.imavis.2020.103911 10.1109/CVPR42600.2020.01164 10.1109/CVPR42600.2020.01189 10.1109/ICCV48922.2021.00272 10.1109/ICCV.2019.00204 10.1109/ICCV48922.2021.00274 10.1109/TITS.2021.3118698 10.1007/978-3-031-20080-9_3 10.1109/CVPR.2018.00745 10.1109/TITS.2022.3177615 10.1109/CVPR46437.2021.00746 10.3390/s18103337 10.1109/TIM.2021.3065438 10.1609/aaai.v35i2.16207 10.1109/CVPR.2016.90 10.1109/CVPR42600.2020.00252 10.1109/CVPR42600.2020.01054 10.1109/CVPR.2019.01298 10.1109/ICCV.2019.00667 10.5244/C.29.150 10.1109/CVPR42600.2020.00466 10.1177/0278364913491297 10.1109/CVPR52688.2022.00535 10.1109/TPAMI.2020.2977026 10.1109/TITS.2019.2892405 10.1109/CVPR46437.2021.01161 10.1109/ICCV.2017.324 10.1109/JSEN.2020.3020626 10.1109/CVPR42600.2020.01011 10.1109/CVPR.2019.00086
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2023
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2023
DBID	97E RIA RIE AAYXX CITATION 7SC 7SP 8FD FR3 JQ2 KR7 L7M L~C L~D
DOI	10.1109/TITS.2023.3273817
DatabaseName	IEEE All-Society Periodicals Package (ASPP) 2005–Present IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef Computer and Information Systems Abstracts Electronics & Communications Abstracts Technology Research Database Engineering Research Database ProQuest Computer Science Collection Civil Engineering Abstracts Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional
DatabaseTitle	CrossRef Civil Engineering Abstracts Technology Research Database Computer and Information Systems Abstracts – Academic Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Engineering Research Database Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Professional
DatabaseTitleList	Civil Engineering Abstracts
Database_xml	– sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering
EISSN	1558-0016
EndPage	10773
ExternalDocumentID	10_1109_TITS_2023_3273817 10128795
Genre	orig-research
GrantInformation_xml	– fundername: National Key Research and Development Program of China grantid: 2022YFB2503302 funderid: 10.13039/501100012166 – fundername: Key Research and Development Program of Jiangsu Province grantid: BE2020083-3 – fundername: National Natural Science Foundation of China grantid: 52225212; U20A20333; 52072160 funderid: 10.13039/501100001809
GroupedDBID	-~X 0R~ 29I 4.4 5GY 5VS 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABQJQ ABVLG ACGFO ACGFS ACIWK ACNCT AENEX AETIX AGQYO AGSQL AHBIQ AIBXA AKJIK AKQYR ALMA_UNASSIGNED_HOLDINGS ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 EBS EJD HZ~ H~9 IFIPE IPLJI JAVBF LAI M43 O9- OCL P2P PQQKQ RIA RIE RNS ZY4 AAYXX CITATION 7SC 7SP 8FD FR3 JQ2 KR7 L7M L~C L~D
ID	FETCH-LOGICAL-c294t-684b5ddf9aaa88a96b8ab297af66e10debaadd461da75692c39a09a0e6616e453
IEDL.DBID	RIE
ISICitedReferencesCount	2
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001005030100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	1524-9050
IngestDate	Sun Nov 09 06:33:56 EST 2025 Tue Nov 18 22:53:28 EST 2025 Sat Nov 29 07:56:38 EST 2025 Wed Aug 27 01:53:19 EDT 2025
IsPeerReviewed	true
IsScholarly	true
Issue	10
Language	English
License	https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c294t-684b5ddf9aaa88a96b8ab297af66e10debaadd461da75692c39a09a0e6616e453
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ORCID	0000-0002-9136-8091 0000-0002-7265-0008 0000-0001-8809-2103 0000-0002-2079-3867 0000-0002-0633-9887
PQID	2872435758
PQPubID	75735
PageCount	14
ParticipantIDs	ieee_primary_10128795 proquest_journals_2872435758 crossref_citationtrail_10_1109_TITS_2023_3273817 crossref_primary_10_1109_TITS_2023_3273817
PublicationCentury	2000
PublicationDate	2023-10-01
PublicationDateYYYYMMDD	2023-10-01
PublicationDate_xml	– month: 10 year: 2023 text: 2023-10-01 day: 01
PublicationDecade	2020
PublicationPlace	New York
PublicationPlace_xml	– name: New York
PublicationTitle	IEEE transactions on intelligent transportation systems
PublicationTitleAbbrev	TITS
PublicationYear	2023
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	ref13 ref35 ref12 ref34 ref15 ref37 ref14 ref31 ref11 ref10 ref32 ge (ref29) 2020 ref2 ref1 loshchilov (ref36) 2017 ref17 ref39 ref16 ref38 ref19 ref18 graham (ref33) 2017 hu (ref30) 2021 ref24 ref23 ref26 ref25 ref20 qi (ref21) 2017; 30 ref22 ren (ref28) 2015; 28 ref27 ref8 ref7 ref9 ref4 ref3 ref5 mao (ref6) 2021
References_xml	– ident: ref11 doi: 10.1109/CVPR.2018.00472 – ident: ref4 doi: 10.1109/TETCI.2023.3235381 – ident: ref12 doi: 10.1016/j.imavis.2020.103911 – ident: ref15 doi: 10.1109/CVPR42600.2020.01164 – ident: ref34 doi: 10.1109/CVPR42600.2020.01189 – ident: ref25 doi: 10.1109/ICCV48922.2021.00272 – ident: ref22 doi: 10.1109/ICCV.2019.00204 – ident: ref19 doi: 10.1109/ICCV48922.2021.00274 – year: 2017 ident: ref36 article-title: Decoupled weight decay regularization publication-title: arXiv 1711 05101 – ident: ref31 doi: 10.1109/TITS.2021.3118698 – ident: ref38 doi: 10.1007/978-3-031-20080-9_3 – ident: ref9 doi: 10.1109/CVPR.2018.00745 – ident: ref1 doi: 10.1109/TITS.2022.3177615 – ident: ref24 doi: 10.1109/CVPR46437.2021.00746 – ident: ref26 doi: 10.3390/s18103337 – ident: ref2 doi: 10.1109/TIM.2021.3065438 – ident: ref17 doi: 10.1609/aaai.v35i2.16207 – volume: 30 start-page: 1 year: 2017 ident: ref21 article-title: PointNet++: Deep hierarchical feature learning on point sets in a metric space publication-title: Proc Adv Neural Inf Process Syst – ident: ref10 doi: 10.1109/CVPR.2016.90 – ident: ref16 doi: 10.1109/CVPR42600.2020.00252 – year: 2021 ident: ref6 article-title: One million scenes for autonomous driving: ONCE dataset publication-title: arXiv 2106 11037 – ident: ref18 doi: 10.1109/CVPR42600.2020.01054 – ident: ref27 doi: 10.1109/CVPR.2019.01298 – year: 2020 ident: ref29 article-title: AFDet: Anchor free one stage 3D object detection publication-title: arXiv 2006 12671 – ident: ref7 doi: 10.1109/ICCV.2019.00667 – ident: ref32 doi: 10.5244/C.29.150 – year: 2021 ident: ref30 article-title: AFDetV2: Rethinking the necessity of the second stage for object detection from point clouds publication-title: arXiv 2112 09205 – ident: ref37 doi: 10.1109/CVPR42600.2020.00466 – ident: ref14 doi: 10.1177/0278364913491297 – volume: 28 start-page: 1 year: 2015 ident: ref28 article-title: Faster R-CNN: Towards real-time object detection with region proposal networks publication-title: Proc Adv Neural Inf Process Syst – ident: ref39 doi: 10.1109/CVPR52688.2022.00535 – ident: ref23 doi: 10.1109/TPAMI.2020.2977026 – ident: ref3 doi: 10.1109/TITS.2019.2892405 – ident: ref8 doi: 10.1109/CVPR46437.2021.01161 – ident: ref35 doi: 10.1109/ICCV.2017.324 – year: 2017 ident: ref33 article-title: Submanifold sparse convolutional networks publication-title: arXiv 1706 01307 – ident: ref5 doi: 10.1109/JSEN.2020.3020626 – ident: ref13 doi: 10.1109/CVPR42600.2020.01011 – ident: ref20 doi: 10.1109/CVPR.2019.00086
SSID	ssj0014511
Score	2.4051924
Snippet	Real-time and accurate 3-D object detection is one of the foundational technologies for environmental perception in autonomous vehicles. However, the existing...
SourceID	proquest crossref ieee
SourceType	Aggregation Database Enrichment Source Index Database Publisher
StartPage	10760
SubjectTerms	3D object detection Accuracy Algorithms anchor-free Autonomous vehicles Computer networks Detection algorithms Feature extraction Location awareness Object detection Object recognition Perception point cloud Point cloud compression single-stage space feature enhancement Training
Title	CenterPoint-SE: A Single-Stage Anchor-Free 3-D Object Detection Algorithm With Spatial Awareness Enhancement
URI	https://ieeexplore.ieee.org/document/10128795 https://www.proquest.com/docview/2872435758
Volume	24
WOSCitedRecordID	wos001005030100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 1558-0016 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014511 issn: 1524-9050 databaseCode: RIE dateStart: 20000101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3dS8MwEA9u-KAPfovTKXnwScjs-pEmvg23oSA62ETfSppeXWF2Uqv---bSTgeiIJRSSNKE_tK7Sy73O0JO09SPjWaXLOBaM-QnYTKFLlMglBKmwKnY9W_C21vx-ChHdbC6jYUBAHv4DDr4aH35yVy_4VbZOXJRYXLsBmmEIa-Ctb5cBki0ZclRXZ-ZHhYuzK4jzyfXk3EH84R3PIxEscnJvpWQzaryQxRb_TLc_OfItshGbUjSXoX8NlmBfIesL9EL7pIZbt1CMZpneWnk5gXt0bEpmAEzJuYT0F5uZF_BhgUA9Vif3sW4KUP7UNrzWTntzZ7mRVZOn-mDuVNMX5xhnx8YQWZEJB3kU5w1OLg9cj8cTC6vWJ1dgWlX-iXjwo-DJEmlQlCU5LFQsStDlXIOXSeBWBnZ5_NuosKAS1d7UjnmAqPROfiBt0-a-TyHA0IFSA8DTTgk2leuUGgYCY8rkeg4kGmLOIvPHemaehwzYMwiuwRxZIQIRYhQVCPUImdfTV4q3o2_Ku8hJEsVKzRapL0ANap_zdfIlLjGRjTrpMNfmh2RNXx7dWSvTZpl8QbHZFW_l9lrcWJn3SfcedPF
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1ba9swFD70Mmj3sHa90Gy96GFPA6WOL4q0N7MmNDTLCs1o3owsH7eB1C6uu_396chOVhgrDIwxSEbCn3zO0eV8H8CnPA9T69kVj4QxnPhJuMqxxzVKraUt8Bp2_XF_MpGzmbpuk9VdLgwiusNn2KVHt5efleaZlsrOiYuKxLHXYZOks9p0rdWmAVFtOXpUP-S2jeUmZs9T59PR9KZLSuHdgHJRnDzZHzfkdFX-MsbOwwx3_rNvu_CuDSVZ3GD_Htaw2IO3LwgG92FBi7dYXZfzoraW8wuL2Y0tWCC3QeYdsriw1q_iwwqRBfyCfU9pWYZdYO1OaBUsXtyV1by-f2C39s5IwHhObf6iHDJrJNmguKdxQ507gB_DwfTrJW_1FbjxVVhzIcM0yrJcaYJFK5FKnfqqr3MhsOdlmGpr_ULRy3Q_Eso3gdKevdD6dIFhFBzCRlEWeARMogoo1URgZkLtS02hkQyElplJI5V3wFt-7sS05OOkgbFI3CTEUwkhlBBCSYtQBz6vXnlsmDdeq3xAkLyo2KDRgeMlqEn7cz4ltsS3UaKdKX34x2tnsHU5_TZOxqPJ1UfYppaaA3zHsFFXz3gCb8zPev5UnboR-BsO8NcO
openUrl	ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=CenterPoint-SE%3A+A+Single-Stage+Anchor-Free+3-D+Object+Detection+Algorithm+With+Spatial+Awareness+Enhancement&rft.jtitle=IEEE+transactions+on+intelligent+transportation+systems&rft.au=Wang%2C+Hai&rft.au=Tao%2C+Le&rft.au=Cai%2C+Yingfeng&rft.au=Chen%2C+Long&rft.date=2023-10-01&rft.issn=1524-9050&rft.eissn=1558-0016&rft.volume=24&rft.issue=10&rft.spage=10760&rft.epage=10773&rft_id=info:doi/10.1109%2FTITS.2023.3273817&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TITS_2023_3273817
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1524-9050&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1524-9050&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1524-9050&client=summon