Please use this identifier to cite or link to this item:
Full metadata record
DC FieldValueLanguage
dc.contributor.CRUESPUniversidade Estadual de Campinaspt
dc.typeArtigo de periódicopt
dc.titleApproximate similarity search for online multimedia services on distributed CPU-GPU platformspt
dc.contributor.authorTeodoro, Gpt
dc.contributor.authorValle, Ept
dc.contributor.authorMariano, Npt
dc.contributor.authorTorres, Rpt
dc.contributor.authorMeira, Wpt
dc.contributor.authorSaltz, JHpt
unicamp.authorTeodoro, George Saltz, Joel H. Emory Univ, Ctr Comprehens Informat, Atlanta, GA 30322 USApt
unicamp.authorValle, Eduardo Univ Estadual Campinas, Recod Lab DCA FEEC, Campinas, SP, Brazilpt
unicamp.authorMariano, Nathan Meira, Wagner, Jr. Univ Fed Minas Gerais, Dept Comp Sci, Belo Horizonte, MG, Brazilpt
unicamp.authorTorres, Ricardo Univ Estadual Campinas, Recod Lab DSI IC, Campinas, SP, Brazilpt
dc.subjectDescriptor indexingpt
dc.subjectMultimedia databasespt
dc.subjectInformation retrievalpt
dc.subject.wosSpace-filling Curvept
dc.subject.wosImage Retrievalpt
dc.description.abstractSimilarity search in high-dimensional spaces is a pivotal operation for several database applications, including online content-based multimedia services. With the increasing popularity of multimedia applications, these services are facing new challenges regarding (1) the very large and growing volumes of data to be indexed/searched and (2) the necessity of reducing the response times as observed by end-users. In addition, the nature of the interactions between users and online services creates fluctuating query request rates throughout execution, which requires a similarity search engine to adapt to better use the computation platform and minimize response times. In this work, we address these challenges with Hypercurves, a flexible framework for answering approximate k-nearest neighbor (kNN) queries for very large multimedia databases. Hypercurves executes in hybrid CPU-GPU environments and is able to attain massive query-processing rates through the cooperative use of these devices. Hypercurves also changes its CPU-GPU task partitioning dynamically according to the observed load, aiming for optimal response times. In our empirical evaluation, dynamic task partitioning reduced query response times by approximately 50 % compared to the best static task partition. Due to a probabilistic proof of equivalence to the sequential kNN algorithm, the CPU-GPU execution of Hypercurves in distributed (multi-node) environments can be aggressively optimized, attaining superlinear scalability while still guaranteeing, with high probability, results at least as good as those from the sequential algorithm.
dc.relation.ispartofVldb Journal
dc.publisher.cityNew Yorkpt
dc.identifier.citationVldb Journal. Springer, v. 23, n. 3, n. 427, n. 448,
dc.sourceWeb of Sciencept
dc.description.sponsorshipConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)pt
dc.description.sponsorshipFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)pt
dc.description.sponsorshipCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)pt
dc.description.sponsorshipFundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG)pt
dc.description.sponsorshipNational Science Foundation [OCI-0910735]pt
dc.description.sponsorship1Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)pt
dc.description.sponsorship1Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)pt
dc.description.sponsorship1Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)pt
dc.description.sponsordocumentnumberCNPq [306580/2012-8, 484254/2012-0]pt
dc.description.sponsordocumentnumberNational Science Foundation [OCI-0910735]pt
dc.description.provenanceMade available in DSpace on 2014-07-30T13:49:25Z (GMT). No. of bitstreams: 0 Previous issue date: 2014en
dc.description.provenanceMade available in DSpace on 2015-11-26T16:44:04Z (GMT). No. of bitstreams: 0 Previous issue date: 2014en
Appears in Collections:Unicamp - Artigos e Outros Documentos

Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.