Label Space-Induced Pseudo Label Refinement for Multi-Source Black-Box Domain Adaptation

Conventional unsupervised domain adaptation (UDA) requires access to source data and/or source model parameters, prohibiting its practical application in terms of privacy, security, and intellectual property. Recent black-box UDA (BDA) reduces such constraints by defining a pseudo label from a singl...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	IEEE transactions on image processing Ročník 34; s. 3181 - 3193
Hlavní autoři:	Yoo, Chaehwa, Liu, Xiaofeng, Xing, Fangxu, Woo, Jonghye, Kang, Je-Won
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	United States IEEE 01.01.2025 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:	Adaptation Adaptation models Application programming interface Application programming interfaces Black boxes black-box Closed box Data models label refinement Labeling Labels multi-source Network security Predictive models Refineries Smoothing methods Sports Training Unsupervised domain adaptation Upper bound
ISSN:	1057-7149, 1941-0042, 1941-0042
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Abstract	Conventional unsupervised domain adaptation (UDA) requires access to source data and/or source model parameters, prohibiting its practical application in terms of privacy, security, and intellectual property. Recent black-box UDA (BDA) reduces such constraints by defining a pseudo label from a single encapsulated source application programming interface (API) prediction, which allows for self-training of the target model. Nonetheless, existing methods have limited consideration for multi-source settings, in which multiple source domain APIs are available to generate pseudo labels. In this work, we introduce a novel training framework for multi-source BDA (MSBDA), dubbed Label Space-Induced Pseudo Label Refinement (LPR). Specifically, LPR incorporates a Pseudo label Refinery Network (PRN) that learns the relationship among source domains conditioned by the target domain only utilizing source API's prediction. The target model is adapted by our dual phases PRN. First, a warm-up phase targets to avoid failure due to noisy samples and provide an initial pseudo-label, which is followed by a label refinement phase with domain relationship exploration. We provide theoretical support for the mechanism of the LPR. Experimental results on four benchmark datasets demonstrate that MSBDA using LPR achieves competitive performance compared to state-of-the-art approaches with different DA settings.
AbstractList	Conventional unsupervised domain adaptation (UDA) requires access to source data and/or source model parameters, prohibiting its practical application in terms of privacy, security, and intellectual property. Recent black-box UDA (BDA) reduces such constraints by defining a pseudo label from a single encapsulated source application programming interface (API) prediction, which allows for self-training of the target model. Nonetheless, existing methods have limited consideration for multi-source settings, in which multiple source domain APIs are available to generate pseudo labels. In this work, we introduce a novel training framework for multi-source BDA (MSBDA), dubbed Label Space-Induced Pseudo Label Refinement (LPR). Specifically, LPR incorporates a Pseudo label Refinery Network (PRN) that learns the relationship among source domains conditioned by the target domain only utilizing source API's prediction. The target model is adapted by our dual phases PRN. First, a warm-up phase targets to avoid failure due to noisy samples and provide an initial pseudo-label, which is followed by a label refinement phase with domain relationship exploration. We provide theoretical support for the mechanism of the LPR. Experimental results on four benchmark datasets demonstrate that MSBDA using LPR achieves competitive performance compared to state-of-the-art approaches with different DA settings.Conventional unsupervised domain adaptation (UDA) requires access to source data and/or source model parameters, prohibiting its practical application in terms of privacy, security, and intellectual property. Recent black-box UDA (BDA) reduces such constraints by defining a pseudo label from a single encapsulated source application programming interface (API) prediction, which allows for self-training of the target model. Nonetheless, existing methods have limited consideration for multi-source settings, in which multiple source domain APIs are available to generate pseudo labels. In this work, we introduce a novel training framework for multi-source BDA (MSBDA), dubbed Label Space-Induced Pseudo Label Refinement (LPR). Specifically, LPR incorporates a Pseudo label Refinery Network (PRN) that learns the relationship among source domains conditioned by the target domain only utilizing source API's prediction. The target model is adapted by our dual phases PRN. First, a warm-up phase targets to avoid failure due to noisy samples and provide an initial pseudo-label, which is followed by a label refinement phase with domain relationship exploration. We provide theoretical support for the mechanism of the LPR. Experimental results on four benchmark datasets demonstrate that MSBDA using LPR achieves competitive performance compared to state-of-the-art approaches with different DA settings. Conventional unsupervised domain adaptation (UDA) requires access to source data and/or source model parameters, prohibiting its practical application in terms of privacy, security, and intellectual property. Recent black-box UDA (BDA) reduces such constraints by defining a pseudo label from a single encapsulated source application programming interface (API) prediction, which allows for self-training of the target model. Nonetheless, existing methods have limited consideration for multi-source settings, in which multiple source domain APIs are available to generate pseudo labels. In this work, we introduce a novel training framework for multi-source BDA (MSBDA), dubbed Label Space-Induced Pseudo Label Refinement (LPR). Specifically, LPR incorporates a Pseudo label Refinery Network (PRN) that learns the relationship among source domains conditioned by the target domain only utilizing source API's prediction. The target model is adapted by our dual phases PRN. First, a warm-up phase targets to avoid failure due to noisy samples and provide an initial pseudo-label, which is followed by a label refinement phase with domain relationship exploration. We provide theoretical support for the mechanism of the LPR. Experimental results on four benchmark datasets demonstrate that MSBDA using LPR achieves competitive performance compared to state-of-the-art approaches with different DA settings.
Author	Yoo, Chaehwa Xing, Fangxu Woo, Jonghye Kang, Je-Won Liu, Xiaofeng
Author_xml	– sequence: 1 givenname: Chaehwa orcidid: 0000-0001-9880-2850 surname: Yoo fullname: Yoo, Chaehwa organization: School of Electrical Engineering, Chungbuk National University, Cheongju, South Korea – sequence: 2 givenname: Xiaofeng orcidid: 0000-0002-4514-2016 surname: Liu fullname: Liu, Xiaofeng organization: Department of Radiology, Gordon Center for Medical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA – sequence: 3 givenname: Fangxu orcidid: 0000-0002-0517-0952 surname: Xing fullname: Xing, Fangxu organization: Department of Radiology, Gordon Center for Medical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA – sequence: 4 givenname: Jonghye orcidid: 0000-0002-5621-9218 surname: Woo fullname: Woo, Jonghye organization: Department of Radiology, Gordon Center for Medical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA – sequence: 5 givenname: Je-Won orcidid: 0000-0002-1637-9479 surname: Kang fullname: Kang, Je-Won email: jewonk@ewha.ac.kr organization: Department of Electronic and Electrical Engineering and the Graduate Program in Smart Factory, Ewha Womans University, Seoul, South Korea
BackLink	https://www.ncbi.nlm.nih.gov/pubmed/40397626$$D View this record in MEDLINE/PubMed
BookMark	eNpdkctrHDEMxk1IaV6991DKQC-5zFZ-r49JmsfCloQmhd4Gjy3DpDP2djwD6X9fL7tpoScJ6fcJSd8JOYwpIiHvKSwoBfP5afWwYMDkgksNjMEBOaZG0BpAsMOSg9S1psIckZOcnwGokFS9JUcCuNGKqWPyY21b7KvHjXVYr6KfHfrqIePsU7VrfcPQRRwwTlVIY_V17qeufkzz6LC67K37WV-ml-pLGmwXqwtvN5OduhTPyJtg-4zv9vGUfL-5frq6q9f3t6uri3XtOPCpVmC1XAbOgnJCCG-0oNTrNijeauOVMS5oF5jxPrBQCtpoC4FiK0VoKfBTcr6buxnTrxnz1Axddtj3NmKac8MZKLk0htKCfvoPfS5nxLJdoag0Uiz1duDHPTW3A_pmM3aDHX83rz8rAOwAN6acRwx_EQrN1pam2NJsbWn2thTJh52kQ8R_OAVYCiX4H8v3hiE
CODEN	IIPRE4
Cites_doi	10.1109/CVPR.2017.316 10.1109/CVPR.2017.18 10.7551/mitpress/7503.003.0045 10.1007/978-3-030-01264-9_9 10.1109/CVPR.2009.5206848 10.1109/CVPR.2019.01053 10.1109/TIP.2021.3124674 10.1109/CVPR.2016.90 10.1117/12.2607895 10.1109/CVPR46437.2021.00997 10.1145/3240508.3240512 10.1109/ICCV.2019.00149 10.1007/978-3-030-58598-3_43 10.1109/CVPR.2017.572 10.1109/CVPR52688.2022.00784 10.1109/CVPR.2018.00887 10.1109/TIP.2019.2950768 10.1109/TPAMI.2021.3103390 10.1109/CVPR.2019.00234 10.1007/978-3-030-01219-9_18 10.1109/CVPR42600.2020.00874 10.1109/CVPR.2018.00417 10.1145/3109859.3109861 10.1109/CVPR.2018.00400 10.1007/978-3-319-49409-8_35 10.1109/CVPR.2017.107 10.1109/ICCV.2013.274 10.1109/TIP.2021.3056212 10.1609/aaai.v32i1.11767 10.1109/CVPR.2018.00392 10.1561/116.00000192 10.24963/ijcai.2021/402 10.1609/aaai.v34i04.6123 10.1609/aaai.v33i01.33015989 10.5555/2946645.2946704 10.1109/CVPR42600.2020.00408 10.1109/TIP.2023.3295929 10.1007/978-3-642-15561-1_16 10.1109/LSP.2022.3194414 10.1109/CVPR42600.2020.01237 10.1109/TIP.2024.3353539 10.1109/CVPR.2017.547 10.1109/TIP.2021.3065254 10.1109/TIP.2023.3258753 10.1109/TIP.2022.3152052
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2025
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2025
DBID	97E RIA RIE AAYXX CITATION NPM 7SC 7SP 8FD JQ2 L7M L~C L~D 7X8
DOI	10.1109/TIP.2025.3570220
DatabaseName	IEEE All-Society Periodicals Package (ASPP) 2005–Present IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef PubMed 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 MEDLINE - Academic
DatabaseTitle	CrossRef PubMed 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 MEDLINE - Academic
DatabaseTitleList	MEDLINE - Academic Technology Research Database PubMed
Database_xml	– sequence: 1 dbid: NPM name: PubMed url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed sourceTypes: Index Database – sequence: 2 dbid: RIE name: IEEE/IET Electronic Library (IEL) (UW System Shared) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher – sequence: 3 dbid: 7X8 name: MEDLINE - Academic url: https://search.proquest.com/medline sourceTypes: Aggregation Database
DeliveryMethod	fulltext_linktorsrc
Discipline	Applied Sciences Engineering
EISSN	1941-0042
EndPage	3193
ExternalDocumentID	40397626 10_1109_TIP_2025_3570220 11008464
Genre	orig-research Journal Article
GrantInformation_xml	– fundername: MSIT (60%) grantid: NRF-2022R1A2C4002052 – fundername: Ministry of Science and ICT (MSIT), South Korea, through the Information Technology Research Center (ITRC) Support Program supervised by the Institute for Information and Communications Technology Planning and Evaluation (IITP) grantid: IITP-2025-RS-2020-II201460 – fundername: Culture, Sports and Tourism Research and Development Program through Korea Creative Content Agency grant funded by the Ministry of Culture, Sports and Tourism in 2024 (20%) grantid: RS-2024-00439534
GroupedDBID	--- -~X .DC 0R~ 29I 4.4 53G 5GY 5VS 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABFSI 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 E.L EBS EJD F5P HZ~ H~9 ICLAB IFIPE IFJZH IPLJI JAVBF LAI M43 MS~ O9- OCL P2P RIA RIE RNS TAE TN5 VH1 AAYXX CITATION NPM RIG 7SC 7SP 8FD JQ2 L7M L~C L~D 7X8
ID	FETCH-LOGICAL-c303t-60a758f32f6c444d97411d7bf63b79d699cf7cf29ddf2f9d6797a0f1eb54fb103
IEDL.DBID	RIE
ISICitedReferencesCount	0
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001502241500001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	1057-7149 1941-0042
IngestDate	Thu Oct 02 22:49:49 EDT 2025 Tue Jul 22 15:12:10 EDT 2025 Mon Jul 21 05:31:03 EDT 2025 Sat Nov 29 07:47:17 EST 2025 Wed Aug 27 01:52:22 EDT 2025
IsPeerReviewed	true
IsScholarly	true
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-c303t-60a758f32f6c444d97411d7bf63b79d699cf7cf29ddf2f9d6797a0f1eb54fb103
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 content type line 23
ORCID	0000-0001-9880-2850 0000-0002-1637-9479 0000-0002-5621-9218 0000-0002-4514-2016 0000-0002-0517-0952
PMID	40397626
PQID	3215954870
PQPubID	85429
PageCount	13
ParticipantIDs	proquest_journals_3215954870 ieee_primary_11008464 crossref_primary_10_1109_TIP_2025_3570220 proquest_miscellaneous_3206589911 pubmed_primary_40397626
PublicationCentury	2000
PublicationDate	2025-01-01
PublicationDateYYYYMMDD	2025-01-01
PublicationDate_xml	– month: 01 year: 2025 text: 2025-01-01 day: 01
PublicationDecade	2020
PublicationPlace	United States
PublicationPlace_xml	– name: United States – name: New York
PublicationTitle	IEEE transactions on image processing
PublicationTitleAbbrev	TIP
PublicationTitleAlternate	IEEE Trans Image Process
PublicationYear	2025
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	ref12 Venkat (ref45); 33 ref56 ref15 ref59 ref58 ref53 ref52 ref11 ref10 ref54 Long (ref28); 37 ref17 ref16 ref18 Hoffman (ref14) Long (ref29) ref51 Liang (ref21) ref46 ref48 ref47 ref42 ref41 ref44 ref43 Wei (ref49) 2020 ref7 ref4 ref3 ref6 ref5 ref40 Dong (ref8); 34 ref35 ref37 Lee (ref19); 3 ref36 ref31 ref33 Zhang (ref57) 2021 ref2 ref1 ref39 ref38 Zhao (ref60); 31 Xie (ref50) Yang (ref55) 2022 ref24 ref23 ref26 ref25 Paszke (ref34) 2017 ref63 ref22 ref27 Dosovitskiy (ref9) Li (ref20); 31 Hoffman (ref13) ref62 Mansour (ref32); 21 ref61 Long (ref30)
References_xml	– ident: ref44 doi: 10.1109/CVPR.2017.316 – ident: ref2 doi: 10.1109/CVPR.2017.18 – ident: ref5 doi: 10.7551/mitpress/7503.003.0045 – start-page: 1 volume-title: Proc. Adv. Neural Inf. Process. Syst. ident: ref13 article-title: Algorithms and theory for multiple-source adaptation – ident: ref3 doi: 10.1007/978-3-030-01264-9_9 – ident: ref6 doi: 10.1109/CVPR.2009.5206848 – ident: ref18 doi: 10.1109/CVPR.2019.01053 – ident: ref15 doi: 10.1109/TIP.2021.3124674 – volume: 3 start-page: 896 issue: 2 volume-title: Proc. Int. Conf. Mach. Learn. (ICML) ident: ref19 article-title: Pseudo-label: The simple and efficient semi-supervised learning method for deep neural networks – ident: ref12 doi: 10.1109/CVPR.2016.90 – ident: ref25 doi: 10.1117/12.2607895 – ident: ref1 doi: 10.1109/CVPR46437.2021.00997 – year: 2022 ident: ref55 article-title: Divide to adapt: Mitigating confirmation bias for domain adaptation of black-box predictors publication-title: arXiv:2205.14467 – volume: 31 start-page: 6799 volume-title: Proc. Adv. Neural Inf. Process. Syst. ident: ref20 article-title: Extracting relationships by multi-domain matching – ident: ref48 doi: 10.1145/3240508.3240512 – volume: 37 start-page: 97 volume-title: Proc. 32nd Int. Conf. Mach. Learn. ident: ref28 article-title: Learning transferable features with deep adaptation networks – start-page: 5423 volume-title: Proc. Int. Conf. Mach. Learn. ident: ref50 article-title: Learning semantic representations for unsupervised domain adaptation – start-page: 1 volume-title: Proc. Int. Conf. Learn. Represent. ident: ref9 article-title: An image is worth 16×16 words: Transformers for image recognition at scale – ident: ref37 doi: 10.1109/ICCV.2019.00149 – ident: ref47 doi: 10.1007/978-3-030-58598-3_43 – volume: 31 start-page: 8559 volume-title: Proc. Adv. Neural Inf. Process. Syst. ident: ref60 article-title: Adversarial multiple source domain adaptation – ident: ref46 doi: 10.1109/CVPR.2017.572 – start-page: 1989 volume-title: Proc. Int. Conf. Mach. Learn. ident: ref14 article-title: CyCADA: Cycle-consistent adversarial domain adaptation – ident: ref23 doi: 10.1109/CVPR52688.2022.00784 – ident: ref42 doi: 10.1109/CVPR.2018.00887 – ident: ref4 doi: 10.1109/TIP.2019.2950768 – volume: 33 start-page: 4647 volume-title: Proc. Adv. Neural Inf. Process. Syst. ident: ref45 article-title: Your classifier can secretly suffice multi-source domain adaptation – volume: 34 start-page: 2848 volume-title: Proc. Adv. Neural Inf. Process. Syst. ident: ref8 article-title: Confident anchor-induced multi-source free domain adaptation – ident: ref22 doi: 10.1109/TPAMI.2021.3103390 – ident: ref33 doi: 10.1109/CVPR.2019.00234 – ident: ref62 doi: 10.1007/978-3-030-01219-9_18 – ident: ref56 doi: 10.1109/CVPR42600.2020.00874 – ident: ref53 doi: 10.1109/CVPR.2018.00417 – ident: ref16 doi: 10.1145/3109859.3109861 – ident: ref59 doi: 10.1109/CVPR.2018.00400 – ident: ref43 doi: 10.1007/978-3-319-49409-8_35 – ident: ref54 doi: 10.1109/CVPR.2017.107 – ident: ref27 doi: 10.1109/ICCV.2013.274 – ident: ref10 doi: 10.1109/TIP.2021.3056212 – year: 2020 ident: ref49 article-title: Theoretical analysis of self-training with deep networks on unlabeled data publication-title: arXiv:2010.03622 – ident: ref36 doi: 10.1609/aaai.v32i1.11767 – ident: ref41 doi: 10.1109/CVPR.2018.00392 – ident: ref26 doi: 10.1561/116.00000192 – ident: ref38 doi: 10.24963/ijcai.2021/402 – year: 2017 ident: ref34 article-title: Automatic differentiation in PyTorch – ident: ref52 doi: 10.1609/aaai.v34i04.6123 – ident: ref61 doi: 10.1609/aaai.v33i01.33015989 – ident: ref11 doi: 10.5555/2946645.2946704 – ident: ref7 doi: 10.1109/CVPR42600.2020.00408 – ident: ref31 doi: 10.1109/TIP.2023.3295929 – ident: ref40 doi: 10.1007/978-3-642-15561-1_16 – volume: 21 start-page: 1 volume-title: Proc. Adv. Neural Inf. Process. Syst. ident: ref32 article-title: Domain adaptation with multiple sources – year: 2021 ident: ref57 article-title: Unsupervised domain adaptation of black-box source models publication-title: arXiv:2101.02839 – ident: ref24 doi: 10.1109/LSP.2022.3194414 – ident: ref51 doi: 10.1109/CVPR42600.2020.01237 – ident: ref17 doi: 10.1109/TIP.2024.3353539 – ident: ref58 doi: 10.1109/CVPR.2017.547 – ident: ref63 doi: 10.1109/TIP.2021.3065254 – ident: ref35 doi: 10.1109/TIP.2023.3258753 – start-page: 2208 volume-title: Proc. Int. Conf. Mach. Learn. ident: ref29 article-title: Deep transfer learning with joint adaptation networks – start-page: 1 volume-title: Proc. Adv. Neural Inf. Process. Syst. ident: ref30 article-title: Conditional adversarial domain adaptation – start-page: 6028 volume-title: Proc. Int. Conf. Mach. Learn. ident: ref21 article-title: Do we really need to access the source data? Source hypothesis transfer for unsupervised domain adaptation – ident: ref39 doi: 10.1109/TIP.2022.3152052
SSID	ssj0014516
Score	2.467999
Snippet	Conventional unsupervised domain adaptation (UDA) requires access to source data and/or source model parameters, prohibiting its practical application in terms...
SourceID	proquest pubmed crossref ieee
SourceType	Aggregation Database Index Database Publisher
StartPage	3181
SubjectTerms	Adaptation Adaptation models Application programming interface Application programming interfaces Black boxes black-box Closed box Data models label refinement Labeling Labels multi-source Network security Predictive models Refineries Smoothing methods Sports Training Unsupervised domain adaptation Upper bound
Title	Label Space-Induced Pseudo Label Refinement for Multi-Source Black-Box Domain Adaptation
URI	https://ieeexplore.ieee.org/document/11008464 https://www.ncbi.nlm.nih.gov/pubmed/40397626 https://www.proquest.com/docview/3215954870 https://www.proquest.com/docview/3206589911
Volume	34
WOSCitedRecordID	wos001502241500001&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/IET Electronic Library (IEL) (UW System Shared) customDbUrl: eissn: 1941-0042 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014516 issn: 1057-7149 databaseCode: RIE dateStart: 19920101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LT-QwDLYWtEJw4P0YFlCQuHAI9JHWkyP7QCAhNOIhza1Km0QaCVrEzCB-PnbameXCgVvVOG0UO_Vnu7YBThS6yJEilCpLUCqrcmnKXEluH4Jlaqu-N6HZBN7e9odDPeiS1UMujHMu_HzmzvgyxPJtU03ZVXbO5c1IX6oFWEDENllrHjLgjrMhtJmhRML9s5hkpM8frgdkCSbZWZohZ5Yuw5KKWBFzSYVP6ij0V_kaagaVc7n2zcWuw2qHLcVFKwwb8MPVm7DW4UzRneLxJqx8KkK4BcMbU7oncU_Gs5PcyaMi4sHYTW0j2qE754mcXycI4oqQsyvvg9dfBAeg_N28i7_NsxnV4sKalza8vw2Pl_8e_lzJrt-CrEiRTWQeGbIefJr4vFJKWTI14thi6fO0RG1zrSuPlU-0tT7xdAM1msjHrsyUL-Mo3YHFuqndHgg0WU7EaRmbRNlYG60ymu1zQ18YG6U9OJ1te_HSltUogjkS6YK4VTC3io5bPdjm3f1P121sDw5mjCq6gzcuUoIwXMMOadrxfJiODMdBTO2aKdMw7iJgHPdgt2Xw_OEzudj_4qW_YJnX1jphDmBx8jp1h_CzepuMxq9HJJfD_lGQyw_mm9sC
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3db9MwED_BQLA9MOgK6xhgJF54cJsPJ54ft0HViVJVrEh9i5zYlipBMq0N2p-_Oyfp9tIH3qL4HFs-O_e7O98dwBchbWBREHKRRJILI1Ku81RwKh8i89gUZ077YhNyNjtbLtW8DVb3sTDWWn_5zA7p0fvyTVXUZCobUXozlJfiKTxLhIjCJlxr6zSgmrPeuZlILhH5d17JQI0WV3PUBaNkGCeSYkv34YUISBRTUoVHAslXWNkNNr3QGR_-53Rfw6sWXbLzZju8gSe27MFhizRZe47XPTh4lIbwCJZTnds_7BrVZ8uplkeBxPO1rU3FmqZf1iE5DccQ5DIftcuvvd2feRMgv6ju2Lfqr16V7Nzom8bB34ff4--LywlvKy7wAkXZhqeBRv3BxZFLCyGEQWUjDI3MXRrnUplUqcLJwkXKGBc5fCGV1IELbZ4Il4dB_Bb2yqq0x8CkTlIkjvNQR8KESiuRYG-XavzHmCAewNdu2bObJrFG5hWSQGXIrYy4lbXcGkCfVveBrl3YAZx2jMrao7fOYgQxlMVOYrfP22Y8NOQJ0aWtaqIh5IXQOBzAu4bB2493--Jkx6Cf4OVk8XOaTa9mP97DPs2zMcmcwt7mtrYf4Hnxb7Na3370u_MeSundYQ
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=Label+Space-Induced+Pseudo+Label+Refinement+for+Multi-Source+Black-Box+Domain+Adaptation&rft.jtitle=IEEE+transactions+on+image+processing&rft.au=Yoo%2C+Chaehwa&rft.au=Liu%2C+Xiaofeng&rft.au=Xing%2C+Fangxu&rft.au=Woo%2C+Jonghye&rft.date=2025-01-01&rft.eissn=1941-0042&rft.volume=34&rft.spage=3181&rft_id=info:doi/10.1109%2FTIP.2025.3570220&rft_id=info%3Apmid%2F40397626&rft.externalDocID=40397626
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1057-7149&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1057-7149&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1057-7149&client=summon