Data mining applied to forensic speaker identification

Autores
Univaso, Pedro Nicolas; Ale, Juan Maria; Gurlekian, Jorge Alberto
Año de publicación
2015
Idioma
español castellano
Tipo de recurso
artículo
Estado
versión publicada
Descripción
In this paper we analyze the advantages of using data mining techniques and tools for data fusion in forensic speaker recognition. Segmental and suprasegmental features were employed in 28 different classifiers, in order to compare their performances. The selected classifiers have different learning techniques: lazy or instance-based, eager and ensemble. Two approaches were employed on the classification task: the use of all features and the use of a feature subset, selected with a gain ratio methodology. The best performances, with all features, were obtained by three classifiers: Logistic Model Tree (eager), LogitBoost (ensemble) and Multilayer Perceptron (eager). Support Vector Machine (eager) proved to be a good classifier if a Pearson VII function-based universal kernel was used. When low dimensional features were selected, ensemble classifiers exceeded the performance of all others classifiers. Segmental and tone features demonstrated the best speaker discrimination capabilities, followed by duration and quality voice features. Evaluation was performed on Argentine-Spanish voice samples from the Speech_Dat database recorded on a fixed telephone environment. Different recording sessions and channels for the test segments were added and the Z-norm procedure was applied for channel compensation.
Fil: Univaso, Pedro Nicolas. Consejo Nacional de Investigaciones Científicas y Técnicas. Oficina de Coordinación Administrativa Houssay. Instituto de Inmunología, Genética y Metabolismo. Universidad de Buenos Aires. Facultad de Medicina. Instituto de Inmunología, Genética y Metabolismo; Argentina
Fil: Ale, Juan Maria. Universidad de Buenos Aires. Facultad de Ingeniería; Argentina
Fil: Gurlekian, Jorge Alberto. Consejo Nacional de Investigaciones Científicas y Técnicas. Oficina de Coordinación Administrativa Houssay. Instituto de Inmunología, Genética y Metabolismo. Universidad de Buenos Aires. Facultad de Medicina. Instituto de Inmunología, Genética y Metabolismo; Argentina
Materia
CLASSIFIERS
DATA FUSION
DATA MINING
ENSEMBLE METHODS
SPEAKER RECOGNITION
Nivel de accesibilidad
acceso abierto
Condiciones de uso
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
Repositorio
CONICET Digital (CONICET)
Institución
Consejo Nacional de Investigaciones Científicas y Técnicas
OAI Identificador
oai:ri.conicet.gov.ar:11336/113286
Acceder
id CONICETDig_e45ba6eb7c4a7188eaf2ff4b251450a0
oai_identifier_str oai:ri.conicet.gov.ar:11336/113286
network_acronym_str CONICETDig
repository_id_str 3498
network_name_str CONICET Digital (CONICET)
spelling Data mining applied to forensic speaker identificationUnivaso, Pedro NicolasAle, Juan MariaGurlekian, Jorge AlbertoCLASSIFIERSDATA FUSIONDATA MININGENSEMBLE METHODSSPEAKER RECOGNITIONhttps://purl.org/becyt/ford/1.2https://purl.org/becyt/ford/1In this paper we analyze the advantages of using data mining techniques and tools for data fusion in forensic speaker recognition. Segmental and suprasegmental features were employed in 28 different classifiers, in order to compare their performances. The selected classifiers have different learning techniques: lazy or instance-based, eager and ensemble. Two approaches were employed on the classification task: the use of all features and the use of a feature subset, selected with a gain ratio methodology. The best performances, with all features, were obtained by three classifiers: Logistic Model Tree (eager), LogitBoost (ensemble) and Multilayer Perceptron (eager). Support Vector Machine (eager) proved to be a good classifier if a Pearson VII function-based universal kernel was used. When low dimensional features were selected, ensemble classifiers exceeded the performance of all others classifiers. Segmental and tone features demonstrated the best speaker discrimination capabilities, followed by duration and quality voice features. Evaluation was performed on Argentine-Spanish voice samples from the Speech_Dat database recorded on a fixed telephone environment. Different recording sessions and channels for the test segments were added and the Z-norm procedure was applied for channel compensation.Fil: Univaso, Pedro Nicolas. Consejo Nacional de Investigaciones Científicas y Técnicas. Oficina de Coordinación Administrativa Houssay. Instituto de Inmunología, Genética y Metabolismo. Universidad de Buenos Aires. Facultad de Medicina. Instituto de Inmunología, Genética y Metabolismo; ArgentinaFil: Ale, Juan Maria. Universidad de Buenos Aires. Facultad de Ingeniería; ArgentinaFil: Gurlekian, Jorge Alberto. Consejo Nacional de Investigaciones Científicas y Técnicas. Oficina de Coordinación Administrativa Houssay. Instituto de Inmunología, Genética y Metabolismo. Universidad de Buenos Aires. Facultad de Medicina. Instituto de Inmunología, Genética y Metabolismo; ArgentinaInstitute of Electrical and Electronics Engineers2015-04-13info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionhttp://purl.org/coar/resource_type/c_6501info:ar-repo/semantics/articuloapplication/pdfapplication/pdfhttp://hdl.handle.net/11336/113286Univaso, Pedro Nicolas; Ale, Juan Maria; Gurlekian, Jorge Alberto; Data mining applied to forensic speaker identification; Institute of Electrical and Electronics Engineers; IEEE Latin America Transactions; 13; 4; 13-4-2015; 1098-11111548-0992CONICET DigitalCONICETspainfo:eu-repo/semantics/altIdentifier/url/https://ieeexplore.ieee.org/document/7106363info:eu-repo/semantics/altIdentifier/doi/10.1109/TLA.2015.7106363info:eu-repo/semantics/openAccesshttps://creativecommons.org/licenses/by-nc-sa/2.5/ar/reponame:CONICET Digital (CONICET)instname:Consejo Nacional de Investigaciones Científicas y Técnicas2025-09-29T09:46:48Zoai:ri.conicet.gov.ar:11336/113286instacron:CONICETInstitucionalhttp://ri.conicet.gov.ar/Organismo científico-tecnológicoNo correspondehttp://ri.conicet.gov.ar/oai/requestdasensio@conicet.gov.ar; lcarlino@conicet.gov.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:34982025-09-29 09:46:48.969CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicasfalse
dc.title.none.fl_str_mv Data mining applied to forensic speaker identification
title Data mining applied to forensic speaker identification
spellingShingle Data mining applied to forensic speaker identification
Univaso, Pedro Nicolas
CLASSIFIERS
DATA FUSION
DATA MINING
ENSEMBLE METHODS
SPEAKER RECOGNITION
title_short Data mining applied to forensic speaker identification
title_full Data mining applied to forensic speaker identification
title_fullStr Data mining applied to forensic speaker identification
title_full_unstemmed Data mining applied to forensic speaker identification
title_sort Data mining applied to forensic speaker identification
dc.creator.none.fl_str_mv Univaso, Pedro Nicolas
Ale, Juan Maria
Gurlekian, Jorge Alberto
author Univaso, Pedro Nicolas
author_facet Univaso, Pedro Nicolas
Ale, Juan Maria
Gurlekian, Jorge Alberto
author_role author
author2 Ale, Juan Maria
Gurlekian, Jorge Alberto
author2_role author
author
dc.subject.none.fl_str_mv CLASSIFIERS
DATA FUSION
DATA MINING
ENSEMBLE METHODS
SPEAKER RECOGNITION
topic CLASSIFIERS
DATA FUSION
DATA MINING
ENSEMBLE METHODS
SPEAKER RECOGNITION
purl_subject.fl_str_mv https://purl.org/becyt/ford/1.2
https://purl.org/becyt/ford/1
dc.description.none.fl_txt_mv In this paper we analyze the advantages of using data mining techniques and tools for data fusion in forensic speaker recognition. Segmental and suprasegmental features were employed in 28 different classifiers, in order to compare their performances. The selected classifiers have different learning techniques: lazy or instance-based, eager and ensemble. Two approaches were employed on the classification task: the use of all features and the use of a feature subset, selected with a gain ratio methodology. The best performances, with all features, were obtained by three classifiers: Logistic Model Tree (eager), LogitBoost (ensemble) and Multilayer Perceptron (eager). Support Vector Machine (eager) proved to be a good classifier if a Pearson VII function-based universal kernel was used. When low dimensional features were selected, ensemble classifiers exceeded the performance of all others classifiers. Segmental and tone features demonstrated the best speaker discrimination capabilities, followed by duration and quality voice features. Evaluation was performed on Argentine-Spanish voice samples from the Speech_Dat database recorded on a fixed telephone environment. Different recording sessions and channels for the test segments were added and the Z-norm procedure was applied for channel compensation.
Fil: Univaso, Pedro Nicolas. Consejo Nacional de Investigaciones Científicas y Técnicas. Oficina de Coordinación Administrativa Houssay. Instituto de Inmunología, Genética y Metabolismo. Universidad de Buenos Aires. Facultad de Medicina. Instituto de Inmunología, Genética y Metabolismo; Argentina
Fil: Ale, Juan Maria. Universidad de Buenos Aires. Facultad de Ingeniería; Argentina
Fil: Gurlekian, Jorge Alberto. Consejo Nacional de Investigaciones Científicas y Técnicas. Oficina de Coordinación Administrativa Houssay. Instituto de Inmunología, Genética y Metabolismo. Universidad de Buenos Aires. Facultad de Medicina. Instituto de Inmunología, Genética y Metabolismo; Argentina
description In this paper we analyze the advantages of using data mining techniques and tools for data fusion in forensic speaker recognition. Segmental and suprasegmental features were employed in 28 different classifiers, in order to compare their performances. The selected classifiers have different learning techniques: lazy or instance-based, eager and ensemble. Two approaches were employed on the classification task: the use of all features and the use of a feature subset, selected with a gain ratio methodology. The best performances, with all features, were obtained by three classifiers: Logistic Model Tree (eager), LogitBoost (ensemble) and Multilayer Perceptron (eager). Support Vector Machine (eager) proved to be a good classifier if a Pearson VII function-based universal kernel was used. When low dimensional features were selected, ensemble classifiers exceeded the performance of all others classifiers. Segmental and tone features demonstrated the best speaker discrimination capabilities, followed by duration and quality voice features. Evaluation was performed on Argentine-Spanish voice samples from the Speech_Dat database recorded on a fixed telephone environment. Different recording sessions and channels for the test segments were added and the Z-norm procedure was applied for channel compensation.
publishDate 2015
dc.date.none.fl_str_mv 2015-04-13
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
http://purl.org/coar/resource_type/c_6501
info:ar-repo/semantics/articulo
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/11336/113286
Univaso, Pedro Nicolas; Ale, Juan Maria; Gurlekian, Jorge Alberto; Data mining applied to forensic speaker identification; Institute of Electrical and Electronics Engineers; IEEE Latin America Transactions; 13; 4; 13-4-2015; 1098-1111
1548-0992
CONICET Digital
CONICET
url http://hdl.handle.net/11336/113286
identifier_str_mv Univaso, Pedro Nicolas; Ale, Juan Maria; Gurlekian, Jorge Alberto; Data mining applied to forensic speaker identification; Institute of Electrical and Electronics Engineers; IEEE Latin America Transactions; 13; 4; 13-4-2015; 1098-1111
1548-0992
CONICET Digital
CONICET
dc.language.none.fl_str_mv spa
language spa
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/url/https://ieeexplore.ieee.org/document/7106363
info:eu-repo/semantics/altIdentifier/doi/10.1109/TLA.2015.7106363
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
eu_rights_str_mv openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv Institute of Electrical and Electronics Engineers
publisher.none.fl_str_mv Institute of Electrical and Electronics Engineers
dc.source.none.fl_str_mv reponame:CONICET Digital (CONICET)
instname:Consejo Nacional de Investigaciones Científicas y Técnicas
reponame_str CONICET Digital (CONICET)
collection CONICET Digital (CONICET)
instname_str Consejo Nacional de Investigaciones Científicas y Técnicas
repository.name.fl_str_mv CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicas
repository.mail.fl_str_mv dasensio@conicet.gov.ar; lcarlino@conicet.gov.ar
_version_ 1844613460979089408
score 13.070432

Publicaciones similares