Unsupervised manifold learning using Reciprocal kNN Graphs in image re-ranking and rank aggregation tasks

dc.creatorPedronette, DCG
dc.creatorPenatti, OAB
dc.creatorTorres, RD
dc.date2014
dc.dateFEB
dc.date2014-07-30T19:32:26Z
dc.date2015-11-26T17:51:06Z
dc.date2014-07-30T19:32:26Z
dc.date2015-11-26T17:51:06Z
dc.date.accessioned2018-03-29T00:34:27Z
dc.date.available2018-03-29T00:34:27Z
dc.identifierImage And Vision Computing. Elsevier Science Bv, v. 32, n. 2, n. 120, n. 130, 2014.
dc.identifier0262-8856
dc.identifier1872-8138
dc.identifierWOS:000332905300003
dc.identifier10.1016/j.imavis.2013.12.009
dc.identifierhttp://www.repositorio.unicamp.br/jspui/handle/REPOSIP/73428
dc.identifierhttp://repositorio.unicamp.br/jspui/handle/REPOSIP/73428
dc.identifier.urihttp://repositorioslatinoamericanos.uchile.cl/handle/2250/1289885
dc.descriptionCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
dc.descriptionFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
dc.descriptionConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
dc.descriptionIn this paper, we present an unsupervised distance learning approach for improving the effectiveness of image retrieval tasks. We propose a Reciprocal kNN Graph algorithm that considers the relationships among ranked lists in the context of a k-reciprocal neighborhood. The similarity is propagated among neighbors considering the geometry of the dataset manifold. The proposed method can be used both for re-ranking and rank aggregation tasks. Unlike traditional diffusion process methods, which require matrix multiplication operations, our algorithm takes only a subset of ranked lists as input, presenting linear complexity in terms of computational and storage requirements. We conducted a large evaluation protocol involving shape, color, and texture descriptors, various datasets, and comparisons with other post-processing approaches. The re-ranking and rank aggregation algorithms yield better results in terms of effectiveness performance than various state-of-the-art algorithms recently proposed in the literature, achieving bull's eye and MAP scores of 100% on the well-known MPEG-7 shape dataset (C) 2013 Elsevier B.V. All rights reserved.
dc.description32
dc.description2
dc.description120
dc.description130
dc.descriptionAMD
dc.descriptionFAEPEX
dc.descriptionCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
dc.descriptionFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
dc.descriptionConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
dc.descriptionCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
dc.descriptionFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
dc.descriptionConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
dc.languageen
dc.publisherElsevier Science Bv
dc.publisherAmsterdam
dc.publisherHolanda
dc.relationImage And Vision Computing
dc.relationImage Vis. Comput.
dc.rightsfechado
dc.rightshttp://www.elsevier.com/about/open-access/open-access-policies/article-posting-policy
dc.sourceWeb of Science
dc.subjectContent-based image retrieval
dc.subjectRe-ranking
dc.subjectRank aggregation
dc.subjectShape
dc.subjectRetrieval
dc.subjectClassification
dc.subjectRecognition
dc.subjectDistance
dc.titleUnsupervised manifold learning using Reciprocal kNN Graphs in image re-ranking and rank aggregation tasks
dc.typeArtículos de revistas


Este ítem pertenece a la siguiente institución