Forensic speaker verification using ordinary least squares

Página do item simplificado
dc.contributor.authorMachado, Thyago J. [UNESP]
dc.contributor.authorFilho, Jozue Vieira [UNESP]
dc.contributor.authorde Oliveira, Mario A.
dc.contributor.institutionUniversidade Estadual Paulista (Unesp)
dc.contributor.institutionMato Grosso Federal Institute of Technology
dc.date.accessioned2020-12-12T02:27:38Z
dc.date.available2020-12-12T02:27:38Z
dc.date.issued2019-10-02
dc.description.abstractIn Brazil, the recognition of speakers for forensic purposes still relies on a subjectivity-based decision-making process through a results analysis of untrustworthy techniques. Owing to the lack of a voice database, speaker verification is currently applied to samples specifically collected for confrontation. However, speaker comparative analysis via contested discourse requires the collection of an excessive amount of voice samples for a series of individuals. Further, the recognition system must inform who is the most compatible with the contested voice from pre-selected individuals. Accordingly, this paper proposes using a combination of linear predictive coding (LPC) and ordinary least squares (OLS) as a speaker verification tool for forensic analysis. The proposed recognition technique establishes confidence and similarity upon which to base forensic reports, indicating verification of the speaker of the contested discourse. Therefore, in this paper, an accurate, quick, alternative method to help verify the speaker is contributed. After running seven different tests, this study preliminarily achieved a hit rate of 100% considering a limited dataset (Brazilian Portuguese). Furthermore, the developed method extracts a larger number of formants, which are indispensable for statistical comparisons via OLS. The proposed framework is robust at certain levels of noise, for sentences with the suppression of word changes, and with different quality or even meaningful audio time differences.en
dc.description.affiliationCampus of Ilha Solteira São Paulo State University (UNESP)
dc.description.affiliationTelecommunications and Aeronautical Engineering São Paulo State University (UNESP)
dc.description.affiliationAutomation and Control Engineering Mato Grosso Federal Institute of Technology
dc.description.affiliationUnespCampus of Ilha Solteira São Paulo State University (UNESP)
dc.description.affiliationUnespTelecommunications and Aeronautical Engineering São Paulo State University (UNESP)
dc.description.sponsorshipCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
dc.description.sponsorshipIdCAPES: 001
dc.identifierhttp://dx.doi.org/10.3390/s19204385
dc.identifier.citationSensors (Switzerland), v. 19, n. 20, 2019.
dc.identifier.doi10.3390/s19204385
dc.identifier.issn1424-8220
dc.identifier.scopus2-s2.0-85073533938
dc.identifier.urihttp://hdl.handle.net/11449/201239
dc.language.isoeng
dc.relation.ispartofSensors (Switzerland)
dc.sourceScopus
dc.subjectForensic phonetics
dc.subjectForensic speaker comparison
dc.subjectLinear predictive coding (LPC)
dc.subjectOrdinary least squares (OLS)
dc.subjectVoice processing
dc.titleForensic speaker verification using ordinary least squaresen
dc.typeArtigo

Arquivos

Coleções

Artigos - Engenharia Elétrica - FEIS