Weakly Supervised Pedestrian Segmentation for Person Re-Identification
Person re-identification (RelD) is an important problem in intelligent surveillance and public security. Among all the solutions to this problem, existing mask-based methods first use a well-pretrained segmentation model to generate a foreground mask, in order to exclude the background from ReID. Th...
Saved in:
| Published in | IEEE transactions on circuits and systems for video technology Vol. 33; no. 3; p. 1 |
|---|---|
| Main Authors | , , , |
| Format | Journal Article |
| Language | English |
| Published |
New York
IEEE
01.03.2023
The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
| Subjects | |
| Online Access | Get full text |
| ISSN | 1051-8215 1558-2205 |
| DOI | 10.1109/TCSVT.2022.3210476 |
Cover
Loading…
| Abstract | Person re-identification (RelD) is an important problem in intelligent surveillance and public security. Among all the solutions to this problem, existing mask-based methods first use a well-pretrained segmentation model to generate a foreground mask, in order to exclude the background from ReID. Then they perform the RelD task directly on the segmented pedestrian image. However, such a process requires extra datasets with pixel-level semantic labels. In this paper, we propose a Weakly Supervised Pedestrian Segmentation (WSPS) framework to produce the foreground mask directly from the RelD datasets. In contrast, our WSPS only requires image-level subject ID labels. To better utilize the pedestrian mask, we also propose the Image Synthesis Augmentation (ISA) technique to further augment the dataset. Experiments show that the features learned from our proposed framework are robust and discriminative. Compared with the baseline, the mAP of our framework is about 4.4%, 11.7%, and 4.0% higher on three widely used datasets including Market-1501, CUHK03, and MSMT17. The code will be available soon. |
|---|---|
| AbstractList | Person re-identification (RelD) is an important problem in intelligent surveillance and public security. Among all the solutions to this problem, existing mask-based methods first use a well-pretrained segmentation model to generate a foreground mask, in order to exclude the background from ReID. Then they perform the RelD task directly on the segmented pedestrian image. However, such a process requires extra datasets with pixel-level semantic labels. In this paper, we propose a Weakly Supervised Pedestrian Segmentation (WSPS) framework to produce the foreground mask directly from the RelD datasets. In contrast, our WSPS only requires image-level subject ID labels. To better utilize the pedestrian mask, we also propose the Image Synthesis Augmentation (ISA) technique to further augment the dataset. Experiments show that the features learned from our proposed framework are robust and discriminative. Compared with the baseline, the mAP of our framework is about 4.4%, 11.7%, and 4.0% higher on three widely used datasets including Market-1501, CUHK03, and MSMT17. The code will be available soon. |
| Author | Jin, Ziqi Wu, Bizhu Shen, Linlin Xie, Jinheng |
| Author_xml | – sequence: 1 givenname: Ziqi surname: Jin fullname: Jin, Ziqi organization: School of Computer Science and Software Engineering, Computer Vision Institute, Shenzhen University, Shenzhen, China – sequence: 2 givenname: Jinheng surname: Xie fullname: Xie, Jinheng organization: School of Computer Science and Software Engineering, Computer Vision Institute, Shenzhen University, Shenzhen, China – sequence: 3 givenname: Bizhu surname: Wu fullname: Wu, Bizhu organization: School of Computer Science and Software Engineering, Computer Vision Institute, Shenzhen University, Shenzhen, China – sequence: 4 givenname: Linlin orcidid: 0000-0003-1420-0815 surname: Shen fullname: Shen, Linlin organization: School of Computer Science and Software Engineering, Computer Vision Institute, Shenzhen University, Shenzhen, China |
| BookMark | eNp9UMtOwzAQtBBItIUfgEskzinrRxLniCoKlSqBaIGj5cZr5NImxU6R-ve4TcWBA6cdaWZ2Z6dPTuumRkKuKAwphfJ2Ppq9zYcMGBtyRkEU-Qnp0SyTKWOQnUYMGU0lo9k56YewBKBCiqJHxu-oP1e7ZLbdoP92AU3yjAZD652ukxl-rLFudeuaOrGNj5wPEb5gOjGRcNZVB_KCnFm9Cnh5nAPyOr6fjx7T6dPDZHQ3TStWZm3KEPSCguQ5X5RVYcEgN2BZqbmQkkVayhjLMKOlBMFzwfUCNeUF19Zawwfkptu78c3XNsZUy2br63hSsUKKnOdM5lHFOlXlmxA8WrXxbq39TlFQ-77UoS-170sd-4om-cdUue7z1mu3-t963VkdIv7eKkvIci74D4bseqw |
| CODEN | ITCTEM |
| CitedBy_id | crossref_primary_10_1109_TCSVT_2024_3442310 crossref_primary_10_1109_TCSVT_2023_3339167 crossref_primary_10_1109_TCSVT_2024_3454171 crossref_primary_10_1109_TCSVT_2023_3341877 |
| Cites_doi | 10.1109/CVPR.2018.00051 10.1109/TPAMI.2018.2820063 10.1109/CVPR.2018.00129 10.1109/CVPR.2018.00607 10.1109/ICPR48806.2021.9412481 10.1145/1015706.1015720 10.1007/978-3-030-58598-3_17 10.1109/ICME.2019.00127 10.1109/TCSVT.2021.3099943 10.1109/TCSVT.2020.3014167 10.1007/978-3-030-01264-9_25 10.1109/ICCV.2019.00380 10.1109/CVPR.2019.00048 10.1109/CVPRW50498.2020.00307 10.1007/978-3-030-58574-7_29 10.1109/ICIP42928.2021.9506058 10.1109/ICCV.2015.133 10.1109/CVPR.2017.660 10.1109/ICCVW.2017.304 10.1109/ICCV48922.2021.00986 10.1109/TCSVT.2018.2873599 10.1109/CVPR.2018.00839 10.1609/aaai.v34i07.6705 10.1109/CVPR.2017.18 10.1109/ICCV.2017.244 10.1109/ICCV.2017.410 10.1109/TCSVT.2020.3047095 10.1007/978-3-319-24574-4_28 10.1109/CVPR.2019.00231 10.24963/ijcai.2021/136 10.1109/ICCV.2019.00032 10.1109/TCSVT.2021.3118060 10.1109/CVPR.2018.00141 10.1109/CVPR.2014.27 10.1016/j.eswa.2022.116636 10.1109/CVPR.2017.389 10.1109/TCSVT.2020.3037179 10.1109/ICCV.2017.405 10.1109/ICPR48806.2021.9412598 10.1109/CVPRW.2019.00190 10.1109/CVPR.2019.00954 10.1109/TCSVT.2021.3088446 10.1109/TPAMI.2009.167 10.1109/CVPR.2017.241 10.1007/978-3-030-01225-0_29 10.1109/TPAMI.2020.3048039 10.1109/TIP.2018.2874715 10.1109/CVPR42600.2020.01229 10.1007/978-3-030-58604-1_22 10.1007/978-3-319-49409-8_35 10.1016/j.imavis.2021.104330 10.1007/s11263-021-01440-4 10.1109/CVPR.2018.00117 10.1007/s11263-021-01474-8 10.1109/CVPR.2016.90 10.1007/978-3-030-58610-2_9 10.1109/CVPR.2018.00016 10.1109/ICCV.2017.322 10.1109/CVPR.2017.316 10.1109/ICME.2019.00092 10.1109/TCSVT.2018.2866260 10.1007/978-3-030-01225-0_30 10.1109/TCSVT.2019.2962073 10.1109/CVPR.2016.319 10.1109/CVPR46437.2021.01037 10.1109/ICCV.2019.00063 10.1109/ICIP42928.2021.9506733 10.1109/TCSVT.2022.3197844 10.1109/ICIP.2019.8803465 10.1016/j.neunet.2020.01.012 10.1109/CVPR.2017.687 10.1109/ICCV.2019.00069 10.1609/aaai.v34i07.7000 |
| 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 JQ2 L7M L~C L~D |
| DOI | 10.1109/TCSVT.2022.3210476 |
| DatabaseName | IEEE Xplore (IEEE) IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef Computer and Information Systems Abstracts Electronics & Communications Abstracts Technology Research Database ProQuest Computer Science Collection Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional |
| DatabaseTitle | CrossRef Technology Research Database Computer and Information Systems Abstracts – Academic Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Professional |
| DatabaseTitleList | Technology Research Database |
| Database_xml | – sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://proxy.k.utb.cz/login?url=https://ieeexplore.ieee.org/ sourceTypes: Publisher |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Engineering |
| EISSN | 1558-2205 |
| EndPage | 1 |
| ExternalDocumentID | 10_1109_TCSVT_2022_3210476 9905634 |
| Genre | orig-research |
| GrantInformation_xml | – fundername: National Natural Science Foundation of China grantid: 91959108 funderid: 10.13039/501100001809 |
| GroupedDBID | -~X 0R~ 29I 4.4 5GY 5VS 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABQJQ ABVLG ACGFO ACGFS ACIWK AENEX AETIX AGQYO AGSQL AHBIQ AI. AIBXA AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 EBS EJD HZ~ H~9 ICLAB IFIPE IFJZH IPLJI JAVBF LAI M43 O9- OCL P2P RIA RIE RNS RXW TAE TN5 VH1 AAYXX CITATION RIG 7SC 7SP 8FD JQ2 L7M L~C L~D |
| ID | FETCH-LOGICAL-c295t-2e0ab108363b9c7f0de3d0f29a348822e088847d2da88043643abea1373afffd3 |
| IEDL.DBID | RIE |
| ISSN | 1051-8215 |
| IngestDate | Mon Jun 30 10:17:08 EDT 2025 Thu Apr 24 22:55:43 EDT 2025 Tue Jul 01 00:41:19 EDT 2025 Wed Aug 27 02:29:16 EDT 2025 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Issue | 3 |
| 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-c295t-2e0ab108363b9c7f0de3d0f29a348822e088847d2da88043643abea1373afffd3 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 |
| ORCID | 0000-0003-1420-0815 0000-0001-5678-4500 0000-0002-6783-6561 0000-0002-7233-5251 |
| PQID | 2784636286 |
| PQPubID | 85433 |
| PageCount | 1 |
| ParticipantIDs | crossref_citationtrail_10_1109_TCSVT_2022_3210476 crossref_primary_10_1109_TCSVT_2022_3210476 proquest_journals_2784636286 ieee_primary_9905634 |
| ProviderPackageCode | CITATION AAYXX |
| PublicationCentury | 2000 |
| PublicationDate | 2023-03-01 |
| PublicationDateYYYYMMDD | 2023-03-01 |
| PublicationDate_xml | – month: 03 year: 2023 text: 2023-03-01 day: 01 |
| PublicationDecade | 2020 |
| PublicationPlace | New York |
| PublicationPlace_xml | – name: New York |
| PublicationTitle | IEEE transactions on circuits and systems for video technology |
| PublicationTitleAbbrev | TCSVT |
| 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 ref57 ref12 ref56 ref15 ref59 ref14 ref58 ref53 ref52 ref11 ref55 ref10 ref54 ref17 ref16 ref19 ref18 ref51 Hou (ref34); 31 ref50 ref46 ref45 ref48 ref47 ref42 ref41 ref44 ref43 ref49 ref8 ref7 ref9 ref4 ref3 ref6 ref5 Simonyan (ref65) Sharma (ref21) 2021 ref80 ref35 ref79 ref78 ref37 ref36 ref31 ref75 ref30 ref74 ref33 ref77 ref32 ref76 ref2 ref1 ref39 Long (ref40); 1 ref71 ref70 ref73 ref72 ref24 ref68 ref23 ref26 ref25 ref69 ref20 ref64 ref63 ref22 ref66 ref28 ref27 ref29 Everingham (ref38) 2021 ref60 ref62 ref61 Krähenbühl (ref67); 24 |
| References_xml | – volume: 31 start-page: 549 volume-title: Proc. Adv. Neural Inf. Process. Syst. ident: ref34 article-title: Self-erasing network for integral object attention – volume: 24 start-page: 109 volume-title: Proc. Adv. Neural Inf. Process. Syst. ident: ref67 article-title: Efficient inference in fully connected CRFs with Gaussian edge potentials – ident: ref72 doi: 10.1109/CVPR.2018.00051 – ident: ref27 doi: 10.1109/TPAMI.2018.2820063 – ident: ref28 doi: 10.1109/CVPR.2018.00129 – ident: ref14 doi: 10.1109/CVPR.2018.00607 – ident: ref22 doi: 10.1109/ICPR48806.2021.9412481 – ident: ref57 doi: 10.1145/1015706.1015720 – ident: ref9 doi: 10.1007/978-3-030-58598-3_17 – ident: ref15 doi: 10.1109/ICME.2019.00127 – ident: ref47 doi: 10.1109/TCSVT.2021.3099943 – volume-title: The PASCAL Visual Object Classes Challenge 2012 (VOC2012) Results year: 2021 ident: ref38 – ident: ref18 doi: 10.1109/TCSVT.2020.3014167 – ident: ref19 doi: 10.1007/978-3-030-01264-9_25 – ident: ref24 doi: 10.1109/ICCV.2019.00380 – ident: ref11 doi: 10.1109/CVPR.2019.00048 – ident: ref49 doi: 10.1109/CVPRW50498.2020.00307 – ident: ref59 doi: 10.1007/978-3-030-58574-7_29 – ident: ref30 doi: 10.1109/ICIP42928.2021.9506058 – ident: ref29 doi: 10.1109/ICCV.2015.133 – ident: ref51 doi: 10.1109/CVPR.2017.660 – ident: ref74 doi: 10.1109/ICCVW.2017.304 – ident: ref69 doi: 10.1109/ICCV48922.2021.00986 – ident: ref73 doi: 10.1109/TCSVT.2018.2873599 – ident: ref77 doi: 10.1109/CVPR.2018.00839 – ident: ref37 doi: 10.1609/aaai.v34i07.6705 – ident: ref43 doi: 10.1109/CVPR.2017.18 – ident: ref48 doi: 10.1109/ICCV.2017.244 – ident: ref78 doi: 10.1109/ICCV.2017.410 – ident: ref45 doi: 10.1109/TCSVT.2020.3047095 – ident: ref64 doi: 10.1007/978-3-319-24574-4_28 – ident: ref35 doi: 10.1109/CVPR.2019.00231 – ident: ref31 doi: 10.24963/ijcai.2021/136 – ident: ref7 doi: 10.1109/ICCV.2019.00032 – ident: ref46 doi: 10.1109/TCSVT.2021.3118060 – ident: ref60 doi: 10.1109/CVPR.2018.00141 – ident: ref61 doi: 10.1109/CVPR.2014.27 – ident: ref58 doi: 10.1016/j.eswa.2022.116636 – ident: ref62 doi: 10.1109/CVPR.2017.389 – ident: ref17 doi: 10.1109/TCSVT.2020.3037179 – ident: ref26 doi: 10.1109/ICCV.2017.405 – ident: ref76 doi: 10.1109/ICPR48806.2021.9412598 – ident: ref80 doi: 10.1109/CVPRW.2019.00190 – ident: ref10 doi: 10.1109/CVPR.2019.00954 – year: 2021 ident: ref21 article-title: Person re-identification with a locally aware transformer publication-title: arXiv:2106.03720 – ident: ref55 doi: 10.1109/TCSVT.2021.3088446 – ident: ref63 doi: 10.1109/TPAMI.2009.167 – ident: ref42 doi: 10.1109/CVPR.2017.241 – ident: ref68 doi: 10.1007/978-3-030-01225-0_29 – ident: ref53 doi: 10.1109/TPAMI.2020.3048039 – ident: ref25 doi: 10.1109/TIP.2018.2874715 – ident: ref36 doi: 10.1109/CVPR42600.2020.01229 – ident: ref71 doi: 10.1007/978-3-030-58604-1_22 – ident: ref41 doi: 10.1007/978-3-319-49409-8_35 – ident: ref20 doi: 10.1016/j.imavis.2021.104330 – ident: ref1 doi: 10.1007/s11263-021-01440-4 – ident: ref5 doi: 10.1109/CVPR.2018.00117 – ident: ref3 doi: 10.1007/s11263-021-01474-8 – ident: ref66 doi: 10.1109/CVPR.2016.90 – ident: ref12 doi: 10.1007/978-3-030-58610-2_9 – ident: ref44 doi: 10.1109/CVPR.2018.00016 – ident: ref50 doi: 10.1109/ICCV.2017.322 – ident: ref39 doi: 10.1109/CVPR.2017.316 – volume: 1 start-page: 97 volume-title: Proc. 32nd Int. Conf. Mach. Learn. (ICML) ident: ref40 article-title: Learning transferable features with deep adaptation networks – ident: ref6 doi: 10.1109/ICME.2019.00092 – ident: ref16 doi: 10.1109/TCSVT.2018.2866260 – ident: ref75 doi: 10.1007/978-3-030-01225-0_30 – ident: ref54 doi: 10.1109/TCSVT.2019.2962073 – ident: ref32 doi: 10.1109/CVPR.2016.319 – ident: ref70 doi: 10.1109/CVPR46437.2021.01037 – ident: ref13 doi: 10.1109/ICCV.2019.00063 – ident: ref23 doi: 10.1109/ICIP42928.2021.9506733 – ident: ref56 doi: 10.1109/TCSVT.2022.3197844 – ident: ref4 doi: 10.1109/ICIP.2019.8803465 – ident: ref2 doi: 10.1016/j.neunet.2020.01.012 – ident: ref8 doi: 10.1109/CVPR.2017.389 – ident: ref33 doi: 10.1109/CVPR.2017.687 – ident: ref52 doi: 10.1109/ICCV.2019.00069 – start-page: 1 volume-title: Proc. Int. Conf. Learn. Represent. ident: ref65 article-title: Very deep convolutional networks for large-scale image recognition – ident: ref79 doi: 10.1609/aaai.v34i07.7000 |
| SSID | ssj0014847 |
| Score | 2.4300346 |
| Snippet | Person re-identification (RelD) is an important problem in intelligent surveillance and public security. Among all the solutions to this problem, existing... |
| SourceID | proquest crossref ieee |
| SourceType | Aggregation Database Enrichment Source Index Database Publisher |
| StartPage | 1 |
| SubjectTerms | Datasets Feature extraction Image contrast Image segmentation Labels Legged locomotion Lips mask-based augmentation Re-Identification Semantics Task analysis Training weakly supervised segmentation |
| Title | Weakly Supervised Pedestrian Segmentation for Person Re-Identification |
| URI | https://ieeexplore.ieee.org/document/9905634 https://www.proquest.com/docview/2784636286 |
| Volume | 33 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV1LTwIxEJ4gJz34QiOKZg_edGG37bL0aIiEmGCIgHLbdNtZD-BCBA766227jxA1xtsm26ZNpzPzTdtvBuCacOV3JPddKWLlMu0B3E6AgatCGsZcMcWp4TsPHtv9CXuYBtMK3JZcGES0j8-waT7tXb5ayI05Kmtpyxm0KduBHR24ZVyt8saAdWwxMQ0X9GjajxUEGY-3xt3R81iHgoQ0DWOFmfwiW07IVlX5YYqtf-kdwKCYWfasZNbcrOOm_PyWtPG_Uz-E_RxoOnfZzjiCCqbHsLeVfrAGvRcUs_mHM9osjcVYoXKGqNBW8kidEb6-5cSk1NHQ1hlacO48oZuxe5P8uO8EJr37cbfv5nUVXEl4sHYJeiL2TVpqGnMZJp5CqryEcEG1OhP9W4fFLFRECa3djGrQImIUPg2pSJJE0VOoposUz8DhhAjJqcElkkmWdHS4xTHhgkmqhCJ18IuFjmSedNzUvphHNvjweGSFExnhRLlw6nBT9llmKTf-bF0zq122zBe6Do1CnlGulavIXLK2qSHjnv_e6wJ2TTn57I1ZA6rr9w1eatCxjq_sbvsCQwjSpQ |
| linkProvider | IEEE |
| linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV07T8MwED5BGYCBN6JQIAMbpCS209QjqqjKowjR8tgix74wtKQVbQb49dhOGlWAEFuk2LLl8919Z_u7AzghXPlNyX1Xili5THsAtxlg4KqQhjFXTHFq-M7du0bnkV2_BC8LcFZyYRDRPj7Duvm0d_lqJDNzVHauLWfQoGwRlrTfD_ycrVXeGbCmLSemAYMeT3uyGUXG4-f9Vu-pr4NBQuqGs8JMhpE5N2TrqvwwxtbDtNehO5tb_rBkUM-mcV1-fkvb-N_Jb8BaATWdi3xvbMICpluwOpeAcBvazygGww-nl42NzZigcu5Roa3lkTo9fH0rqEmpo8Gtc2_hufOAbs7vTYoDvx14bF_2Wx23qKzgSsKDqUvQE7FvElPTmMsw8RRS5SWEC6oVmujfOjBmoSJKaP1mVMMWEaPwaUhFkiSK7kIlHaW4Bw4nREhODTKRTLKkqQMujgkXTFIlFKmCP1voSBZpx031i2Fkww-PR1Y4kRFOVAinCqdln3GedOPP1ttmtcuWxUJXoTaTZ1To5SQy16wNaui4-7_3OoblTr97G91e3d0cwIopLp-_OKtBZfqe4aGGINP4yO68L2vl1e4 |
| 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=Weakly+Supervised+Pedestrian+Segmentation+for+Person+Re-Identification&rft.jtitle=IEEE+transactions+on+circuits+and+systems+for+video+technology&rft.au=Jin%2C+Ziqi&rft.au=Xie%2C+Jinheng&rft.au=Wu%2C+Bizhu&rft.au=Shen%2C+Linlin&rft.date=2023-03-01&rft.pub=IEEE&rft.issn=1051-8215&rft.spage=1&rft.epage=1&rft_id=info:doi/10.1109%2FTCSVT.2022.3210476&rft.externalDocID=9905634 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1051-8215&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1051-8215&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1051-8215&client=summon |