A multi-level rank correlation measure for image retrieval

Detalhes bibliográficos
Autor(a) principal: De Sá, Nikolas Gomes [UNESP]
Data de Publicação: 2021
Outros Autores: Valem, Lucas Pascotti [UNESP], Guimarães Pedronette, Daniel Carlos [UNESP]
Tipo de documento: Artigo de conferência
Idioma: eng
Título da fonte: Repositório Institucional da UNESP
Texto Completo: http://hdl.handle.net/11449/206095
Resumo: Accurately ranking the most relevant elements in a given scenario often represents a central challenge in many applications, composing the core of retrieval systems. Once ranking structures encode relevant similarity information, measuring how correlated are two rank results represents a fundamental task, with diversified applications. In this work, we propose a new rank correlation measure called Multi-Level Rank Correlation Measure (MLCM), which employs a novel approach based on a multi-level analysis for estimating the correlation between ranked lists. While traditional weighted measures assign more relevance to top positions, our proposed approach goes beyond by considering the position at different levels in the ranked lists. The effectiveness of the proposed measure was assessed in unsupervised and weakly supervised learning tasks for image retrieval. The experimental evaluation considered 6 correlation measures as baselines, 3 different image datasets, and multiple features. The results are competitive or, in most of the cases, superior to the baselines, achieving significant effectiveness gains.
id UNSP_495f2a4564d8fa83c5f40737ea39809d
oai_identifier_str oai:repositorio.unesp.br:11449/206095
network_acronym_str UNSP
network_name_str Repositório Institucional da UNESP
repository_id_str 2946
spelling A multi-level rank correlation measure for image retrievalContent-based image retrievalInformation retrievalRank correlationUnsupervised learningAccurately ranking the most relevant elements in a given scenario often represents a central challenge in many applications, composing the core of retrieval systems. Once ranking structures encode relevant similarity information, measuring how correlated are two rank results represents a fundamental task, with diversified applications. In this work, we propose a new rank correlation measure called Multi-Level Rank Correlation Measure (MLCM), which employs a novel approach based on a multi-level analysis for estimating the correlation between ranked lists. While traditional weighted measures assign more relevance to top positions, our proposed approach goes beyond by considering the position at different levels in the ranked lists. The effectiveness of the proposed measure was assessed in unsupervised and weakly supervised learning tasks for image retrieval. The experimental evaluation considered 6 correlation measures as baselines, 3 different image datasets, and multiple features. The results are competitive or, in most of the cases, superior to the baselines, achieving significant effectiveness gains.Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)Department of Statistics Applied Math. and Computing São Paulo State University (UNESP)Department of Statistics Applied Math. and Computing São Paulo State University (UNESP)FAPESP: #2017/25908-6FAPESP: #2018/15597-6FAPESP: #2019/11104-8FAPESP: #2020/11366-0CNPq: #308194/2017-9Universidade Estadual Paulista (Unesp)De Sá, Nikolas Gomes [UNESP]Valem, Lucas Pascotti [UNESP]Guimarães Pedronette, Daniel Carlos [UNESP]2021-06-25T10:26:25Z2021-06-25T10:26:25Z2021-01-01info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/conferenceObject370-378VISIGRAPP 2021 - Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, v. 5, p. 370-378.http://hdl.handle.net/11449/2060952-s2.0-85102977558Scopusreponame:Repositório Institucional da UNESPinstname:Universidade Estadual Paulista (UNESP)instacron:UNESPengVISIGRAPP 2021 - Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applicationsinfo:eu-repo/semantics/openAccess2021-10-22T20:56:23Zoai:repositorio.unesp.br:11449/206095Repositório InstitucionalPUBhttp://repositorio.unesp.br/oai/requestopendoar:29462024-08-05T14:04:16.016072Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)false
dc.title.none.fl_str_mv A multi-level rank correlation measure for image retrieval
title A multi-level rank correlation measure for image retrieval
spellingShingle A multi-level rank correlation measure for image retrieval
De Sá, Nikolas Gomes [UNESP]
Content-based image retrieval
Information retrieval
Rank correlation
Unsupervised learning
title_short A multi-level rank correlation measure for image retrieval
title_full A multi-level rank correlation measure for image retrieval
title_fullStr A multi-level rank correlation measure for image retrieval
title_full_unstemmed A multi-level rank correlation measure for image retrieval
title_sort A multi-level rank correlation measure for image retrieval
author De Sá, Nikolas Gomes [UNESP]
author_facet De Sá, Nikolas Gomes [UNESP]
Valem, Lucas Pascotti [UNESP]
Guimarães Pedronette, Daniel Carlos [UNESP]
author_role author
author2 Valem, Lucas Pascotti [UNESP]
Guimarães Pedronette, Daniel Carlos [UNESP]
author2_role author
author
dc.contributor.none.fl_str_mv Universidade Estadual Paulista (Unesp)
dc.contributor.author.fl_str_mv De Sá, Nikolas Gomes [UNESP]
Valem, Lucas Pascotti [UNESP]
Guimarães Pedronette, Daniel Carlos [UNESP]
dc.subject.por.fl_str_mv Content-based image retrieval
Information retrieval
Rank correlation
Unsupervised learning
topic Content-based image retrieval
Information retrieval
Rank correlation
Unsupervised learning
description Accurately ranking the most relevant elements in a given scenario often represents a central challenge in many applications, composing the core of retrieval systems. Once ranking structures encode relevant similarity information, measuring how correlated are two rank results represents a fundamental task, with diversified applications. In this work, we propose a new rank correlation measure called Multi-Level Rank Correlation Measure (MLCM), which employs a novel approach based on a multi-level analysis for estimating the correlation between ranked lists. While traditional weighted measures assign more relevance to top positions, our proposed approach goes beyond by considering the position at different levels in the ranked lists. The effectiveness of the proposed measure was assessed in unsupervised and weakly supervised learning tasks for image retrieval. The experimental evaluation considered 6 correlation measures as baselines, 3 different image datasets, and multiple features. The results are competitive or, in most of the cases, superior to the baselines, achieving significant effectiveness gains.
publishDate 2021
dc.date.none.fl_str_mv 2021-06-25T10:26:25Z
2021-06-25T10:26:25Z
2021-01-01
dc.type.status.fl_str_mv info:eu-repo/semantics/publishedVersion
dc.type.driver.fl_str_mv info:eu-repo/semantics/conferenceObject
format conferenceObject
status_str publishedVersion
dc.identifier.uri.fl_str_mv VISIGRAPP 2021 - Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, v. 5, p. 370-378.
http://hdl.handle.net/11449/206095
2-s2.0-85102977558
identifier_str_mv VISIGRAPP 2021 - Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, v. 5, p. 370-378.
2-s2.0-85102977558
url http://hdl.handle.net/11449/206095
dc.language.iso.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv VISIGRAPP 2021 - Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications
dc.rights.driver.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv 370-378
dc.source.none.fl_str_mv Scopus
reponame:Repositório Institucional da UNESP
instname:Universidade Estadual Paulista (UNESP)
instacron:UNESP
instname_str Universidade Estadual Paulista (UNESP)
instacron_str UNESP
institution UNESP
reponame_str Repositório Institucional da UNESP
collection Repositório Institucional da UNESP
repository.name.fl_str_mv Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)
repository.mail.fl_str_mv
_version_ 1808128312078237696