Unsupervised Horizontal Pyramid Similarity Learning for Cross-domain Adaptive Person Re-identification

Although person re-identification has made great progress, unsupervised cross-domain adaptive person re-identification is still a challenging problem. With no labeled data in target domain, the performance may have a significant drop. In this paper, we propose an unsupervised cross-domain adaptive p...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE access 2021-01, Vol.9, p.1-1
Hauptverfasser: Dong, Wenhui, Qu, Peishu, Liu, Chunsheng, Tang, Yanke, Gai, Ning
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 1
container_issue
container_start_page 1
container_title IEEE access
container_volume 9
creator Dong, Wenhui
Qu, Peishu
Liu, Chunsheng
Tang, Yanke
Gai, Ning
description Although person re-identification has made great progress, unsupervised cross-domain adaptive person re-identification is still a challenging problem. With no labeled data in target domain, the performance may have a significant drop. In this paper, we propose an unsupervised cross-domain adaptive person re-identification framework based on horizontal pyramid similarity learning (UHPS). Firstly, horizontal pyramid features are extracted by dividing the deep feature maps into different number of partial feature bins. These feature bins with diverse scales can incorporate not only the global information but also local information in different spatial scales, making the framework more robust in complex environment. Then, horizontal pyramid similarity learning is proposed with the mechanism of fusing together the internal similarity of the target domain and the similarity between the source domain and target domain. Finally, the unsupervised clustering algorithm DBSCAN embeded with the horizontal pyramid similarity is employed to select training data in the target domain and estimate the pseudo labels in each training iteration, for the purpose of adapting the framework to the target domain. The results on Market1501 and DukeMTMC-reID confirm that the proposed framework can adapt to the target domain effectively and outperforms the state-of-the-art unsupervised cross domain person re-identification approaches.
doi_str_mv 10.1109/ACCESS.2021.3093083
format Article
fullrecord <record><control><sourceid>proquest_ieee_</sourceid><recordid>TN_cdi_proquest_journals_2547642363</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9466837</ieee_id><doaj_id>oai_doaj_org_article_00f1a4c570ab4d42843fc762b6bbb1ba</doaj_id><sourcerecordid>2547642363</sourcerecordid><originalsourceid>FETCH-LOGICAL-c408t-12e45b71497513e93e899e60f2025ff50d228889ca58336cbff57d9113a47eeb3</originalsourceid><addsrcrecordid>eNpNUdtqGzEQXUoLDWm-IC-CPq-r--XRLGkTMDTUzbOQVqMgY0uutA64X991NoTOywyHOWcup-tuCV4Rgs239TDcbbcriilZMWwY1uxDd0WJND0TTH78r_7c3bS2w3PoGRLqqotPuZ2OUF9Sg4DuS01_S57cHj2eqzukgLbpkPaupumMNuBqTvkZxVLRUEtrfSgHlzJaB3ec0gugR6itZPQL-hQgTymm0U2p5C_dp-j2DW7e8nX39P3u93Dfb37-eBjWm37kWE89ocCFV4QbJQgDw0AbAxLH-TgRo8CBUq21GZ3QjMnRz5gKhhDmuALw7Lp7WHRDcTt7rOng6tkWl-wrUOqzdXVK4x4sxpE4PgqFneeBU81ZHJWkXnrviXez1tdF61jLnxO0ye7KqeZ5fUsFV5JTJtncxZau8fKQCvF9KsH24o9d_LEXf-ybPzPrdmElAHhnGC6lZor9A8PUjHw</addsrcrecordid><sourcetype>Open Website</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2547642363</pqid></control><display><type>article</type><title>Unsupervised Horizontal Pyramid Similarity Learning for Cross-domain Adaptive Person Re-identification</title><source>IEEE Open Access Journals</source><source>DOAJ Directory of Open Access Journals</source><source>EZB-FREE-00999 freely available EZB journals</source><creator>Dong, Wenhui ; Qu, Peishu ; Liu, Chunsheng ; Tang, Yanke ; Gai, Ning</creator><creatorcontrib>Dong, Wenhui ; Qu, Peishu ; Liu, Chunsheng ; Tang, Yanke ; Gai, Ning</creatorcontrib><description>Although person re-identification has made great progress, unsupervised cross-domain adaptive person re-identification is still a challenging problem. With no labeled data in target domain, the performance may have a significant drop. In this paper, we propose an unsupervised cross-domain adaptive person re-identification framework based on horizontal pyramid similarity learning (UHPS). Firstly, horizontal pyramid features are extracted by dividing the deep feature maps into different number of partial feature bins. These feature bins with diverse scales can incorporate not only the global information but also local information in different spatial scales, making the framework more robust in complex environment. Then, horizontal pyramid similarity learning is proposed with the mechanism of fusing together the internal similarity of the target domain and the similarity between the source domain and target domain. Finally, the unsupervised clustering algorithm DBSCAN embeded with the horizontal pyramid similarity is employed to select training data in the target domain and estimate the pseudo labels in each training iteration, for the purpose of adapting the framework to the target domain. The results on Market1501 and DukeMTMC-reID confirm that the proposed framework can adapt to the target domain effectively and outperforms the state-of-the-art unsupervised cross domain person re-identification approaches.</description><identifier>ISSN: 2169-3536</identifier><identifier>EISSN: 2169-3536</identifier><identifier>DOI: 10.1109/ACCESS.2021.3093083</identifier><identifier>CODEN: IAECCG</identifier><language>eng</language><publisher>Piscataway: IEEE</publisher><subject>Adaptation models ; Algorithms ; Bins ; Clustering ; Data models ; Deep learning ; Domains ; Feature extraction ; Feature maps ; Generative adversarial networks ; Horizontal pyramid similarity learning ; Learning ; Person re-identification ; Similarity ; Surveillance ; Training ; Unsupervised cross domain adaption ; Unsupervised deep learning</subject><ispartof>IEEE access, 2021-01, Vol.9, p.1-1</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c408t-12e45b71497513e93e899e60f2025ff50d228889ca58336cbff57d9113a47eeb3</citedby><cites>FETCH-LOGICAL-c408t-12e45b71497513e93e899e60f2025ff50d228889ca58336cbff57d9113a47eeb3</cites><orcidid>0000-0001-5157-8951 ; 0000-0002-3218-3644 ; 0000-0001-9770-0596 ; 0000-0001-5516-2486 ; 0000-0001-5949-2984</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9466837$$EHTML$$P50$$Gieee$$Hfree_for_read</linktohtml><link.rule.ids>314,780,784,864,2102,27633,27924,27925,54933</link.rule.ids></links><search><creatorcontrib>Dong, Wenhui</creatorcontrib><creatorcontrib>Qu, Peishu</creatorcontrib><creatorcontrib>Liu, Chunsheng</creatorcontrib><creatorcontrib>Tang, Yanke</creatorcontrib><creatorcontrib>Gai, Ning</creatorcontrib><title>Unsupervised Horizontal Pyramid Similarity Learning for Cross-domain Adaptive Person Re-identification</title><title>IEEE access</title><addtitle>Access</addtitle><description>Although person re-identification has made great progress, unsupervised cross-domain adaptive person re-identification is still a challenging problem. With no labeled data in target domain, the performance may have a significant drop. In this paper, we propose an unsupervised cross-domain adaptive person re-identification framework based on horizontal pyramid similarity learning (UHPS). Firstly, horizontal pyramid features are extracted by dividing the deep feature maps into different number of partial feature bins. These feature bins with diverse scales can incorporate not only the global information but also local information in different spatial scales, making the framework more robust in complex environment. Then, horizontal pyramid similarity learning is proposed with the mechanism of fusing together the internal similarity of the target domain and the similarity between the source domain and target domain. Finally, the unsupervised clustering algorithm DBSCAN embeded with the horizontal pyramid similarity is employed to select training data in the target domain and estimate the pseudo labels in each training iteration, for the purpose of adapting the framework to the target domain. The results on Market1501 and DukeMTMC-reID confirm that the proposed framework can adapt to the target domain effectively and outperforms the state-of-the-art unsupervised cross domain person re-identification approaches.</description><subject>Adaptation models</subject><subject>Algorithms</subject><subject>Bins</subject><subject>Clustering</subject><subject>Data models</subject><subject>Deep learning</subject><subject>Domains</subject><subject>Feature extraction</subject><subject>Feature maps</subject><subject>Generative adversarial networks</subject><subject>Horizontal pyramid similarity learning</subject><subject>Learning</subject><subject>Person re-identification</subject><subject>Similarity</subject><subject>Surveillance</subject><subject>Training</subject><subject>Unsupervised cross domain adaption</subject><subject>Unsupervised deep learning</subject><issn>2169-3536</issn><issn>2169-3536</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>ESBDL</sourceid><sourceid>RIE</sourceid><sourceid>DOA</sourceid><recordid>eNpNUdtqGzEQXUoLDWm-IC-CPq-r--XRLGkTMDTUzbOQVqMgY0uutA64X991NoTOywyHOWcup-tuCV4Rgs239TDcbbcriilZMWwY1uxDd0WJND0TTH78r_7c3bS2w3PoGRLqqotPuZ2OUF9Sg4DuS01_S57cHj2eqzukgLbpkPaupumMNuBqTvkZxVLRUEtrfSgHlzJaB3ec0gugR6itZPQL-hQgTymm0U2p5C_dp-j2DW7e8nX39P3u93Dfb37-eBjWm37kWE89ocCFV4QbJQgDw0AbAxLH-TgRo8CBUq21GZ3QjMnRz5gKhhDmuALw7Lp7WHRDcTt7rOng6tkWl-wrUOqzdXVK4x4sxpE4PgqFneeBU81ZHJWkXnrviXez1tdF61jLnxO0ye7KqeZ5fUsFV5JTJtncxZau8fKQCvF9KsH24o9d_LEXf-ybPzPrdmElAHhnGC6lZor9A8PUjHw</recordid><startdate>20210101</startdate><enddate>20210101</enddate><creator>Dong, Wenhui</creator><creator>Qu, Peishu</creator><creator>Liu, Chunsheng</creator><creator>Tang, Yanke</creator><creator>Gai, Ning</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>ESBDL</scope><scope>RIA</scope><scope>RIE</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>7SR</scope><scope>8BQ</scope><scope>8FD</scope><scope>JG9</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><scope>DOA</scope><orcidid>https://orcid.org/0000-0001-5157-8951</orcidid><orcidid>https://orcid.org/0000-0002-3218-3644</orcidid><orcidid>https://orcid.org/0000-0001-9770-0596</orcidid><orcidid>https://orcid.org/0000-0001-5516-2486</orcidid><orcidid>https://orcid.org/0000-0001-5949-2984</orcidid></search><sort><creationdate>20210101</creationdate><title>Unsupervised Horizontal Pyramid Similarity Learning for Cross-domain Adaptive Person Re-identification</title><author>Dong, Wenhui ; Qu, Peishu ; Liu, Chunsheng ; Tang, Yanke ; Gai, Ning</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c408t-12e45b71497513e93e899e60f2025ff50d228889ca58336cbff57d9113a47eeb3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Adaptation models</topic><topic>Algorithms</topic><topic>Bins</topic><topic>Clustering</topic><topic>Data models</topic><topic>Deep learning</topic><topic>Domains</topic><topic>Feature extraction</topic><topic>Feature maps</topic><topic>Generative adversarial networks</topic><topic>Horizontal pyramid similarity learning</topic><topic>Learning</topic><topic>Person re-identification</topic><topic>Similarity</topic><topic>Surveillance</topic><topic>Training</topic><topic>Unsupervised cross domain adaption</topic><topic>Unsupervised deep learning</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Dong, Wenhui</creatorcontrib><creatorcontrib>Qu, Peishu</creatorcontrib><creatorcontrib>Liu, Chunsheng</creatorcontrib><creatorcontrib>Tang, Yanke</creatorcontrib><creatorcontrib>Gai, Ning</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE Open Access Journals</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics &amp; Communications Abstracts</collection><collection>Engineered Materials Abstracts</collection><collection>METADEX</collection><collection>Technology Research Database</collection><collection>Materials Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts – Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><collection>DOAJ Directory of Open Access Journals</collection><jtitle>IEEE access</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Dong, Wenhui</au><au>Qu, Peishu</au><au>Liu, Chunsheng</au><au>Tang, Yanke</au><au>Gai, Ning</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Unsupervised Horizontal Pyramid Similarity Learning for Cross-domain Adaptive Person Re-identification</atitle><jtitle>IEEE access</jtitle><stitle>Access</stitle><date>2021-01-01</date><risdate>2021</risdate><volume>9</volume><spage>1</spage><epage>1</epage><pages>1-1</pages><issn>2169-3536</issn><eissn>2169-3536</eissn><coden>IAECCG</coden><abstract>Although person re-identification has made great progress, unsupervised cross-domain adaptive person re-identification is still a challenging problem. With no labeled data in target domain, the performance may have a significant drop. In this paper, we propose an unsupervised cross-domain adaptive person re-identification framework based on horizontal pyramid similarity learning (UHPS). Firstly, horizontal pyramid features are extracted by dividing the deep feature maps into different number of partial feature bins. These feature bins with diverse scales can incorporate not only the global information but also local information in different spatial scales, making the framework more robust in complex environment. Then, horizontal pyramid similarity learning is proposed with the mechanism of fusing together the internal similarity of the target domain and the similarity between the source domain and target domain. Finally, the unsupervised clustering algorithm DBSCAN embeded with the horizontal pyramid similarity is employed to select training data in the target domain and estimate the pseudo labels in each training iteration, for the purpose of adapting the framework to the target domain. The results on Market1501 and DukeMTMC-reID confirm that the proposed framework can adapt to the target domain effectively and outperforms the state-of-the-art unsupervised cross domain person re-identification approaches.</abstract><cop>Piscataway</cop><pub>IEEE</pub><doi>10.1109/ACCESS.2021.3093083</doi><tpages>1</tpages><orcidid>https://orcid.org/0000-0001-5157-8951</orcidid><orcidid>https://orcid.org/0000-0002-3218-3644</orcidid><orcidid>https://orcid.org/0000-0001-9770-0596</orcidid><orcidid>https://orcid.org/0000-0001-5516-2486</orcidid><orcidid>https://orcid.org/0000-0001-5949-2984</orcidid><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 2169-3536
ispartof IEEE access, 2021-01, Vol.9, p.1-1
issn 2169-3536
2169-3536
language eng
recordid cdi_proquest_journals_2547642363
source IEEE Open Access Journals; DOAJ Directory of Open Access Journals; EZB-FREE-00999 freely available EZB journals
subjects Adaptation models
Algorithms
Bins
Clustering
Data models
Deep learning
Domains
Feature extraction
Feature maps
Generative adversarial networks
Horizontal pyramid similarity learning
Learning
Person re-identification
Similarity
Surveillance
Training
Unsupervised cross domain adaption
Unsupervised deep learning
title Unsupervised Horizontal Pyramid Similarity Learning for Cross-domain Adaptive Person Re-identification
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T17%3A42%3A04IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_ieee_&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Unsupervised%20Horizontal%20Pyramid%20Similarity%20Learning%20for%20Cross-domain%20Adaptive%20Person%20Re-identification&rft.jtitle=IEEE%20access&rft.au=Dong,%20Wenhui&rft.date=2021-01-01&rft.volume=9&rft.spage=1&rft.epage=1&rft.pages=1-1&rft.issn=2169-3536&rft.eissn=2169-3536&rft.coden=IAECCG&rft_id=info:doi/10.1109/ACCESS.2021.3093083&rft_dat=%3Cproquest_ieee_%3E2547642363%3C/proquest_ieee_%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2547642363&rft_id=info:pmid/&rft_ieee_id=9466837&rft_doaj_id=oai_doaj_org_article_00f1a4c570ab4d42843fc762b6bbb1ba&rfr_iscdi=true