Publication:
Comparison of spectrum estimators in speaker verification: Mismatch conditions induced by vocal effort

dc.contributor.authorKinnunen, Tomi
dc.contributor.authorRajan, Padmanabhan
dc.contributor.authorPohjalainen, Jouni
dc.contributor.authorAlku, Paavo
dc.contributor.authorBimbot, F.
dc.contributor.authorCerisara, C.
dc.contributor.authorFougeron, C.
dc.contributor.authorGravier, G.
dc.contributor.authorLamel, L.
dc.contributor.authorPellegrino, F.
dc.contributor.authorPerrier, P.
dc.contributor.buuauthorHanilçi, Cemal
dc.contributor.buuauthorErtaş, Figen
dc.contributor.departmentMühendislik Fakültesi
dc.contributor.departmentElektrik ve Elektronik Mühendisliği Bölümü
dc.contributor.researcheridAAH-4188-2021
dc.contributor.researcheridS-4967-2016
dc.contributor.scopusid35781455400
dc.contributor.scopusid24724154500
dc.date.accessioned2023-05-09T11:34:10Z
dc.date.available2023-05-09T11:34:10Z
dc.date.issued2013
dc.descriptionBu çalışma, 25-29 Ağustos 2013 tarihleri arasında Lyon[Fransa]’da düzenlenen 14. Annual Conference of the International-Speech-Communication-Association (INTERSPEECH 2013)’da bildiri olarak sunulmuştur.
dc.description.abstractWe study the problem of vocal effort mismatch in speaker verification. Changes in speaker's vocal effort induce changes in fundamental frequency (F0) and formant structure which introduce unwanted intra-speaker variations to features. We compare seven alternative spectrum estimators in the context of melfrequency cepstral coefficient (MFCC) extraction for speaker verification. The compared variants include traditional FFT spectrum and six parametric all-pole models. Experimental results on the NIST 2010 speaker recognition evaluation (SRE) corpus utilizing both GMM-UBM and more recent GMM supervector classifier indicate that spectrum estimation has a considerable impact on speaker verification accuracy under mismatched vocal effort conditions. The highest recognition accuracy was achieved using a particular variant of temporally weighted all-pole model, stabilized weighted linear prediction (SWLP).
dc.description.sponsorshipAcademy of Finland (253120)
dc.description.sponsorshipInt Speech Commun Assoc
dc.description.sponsorshipEuropa org
dc.description.sponsorshipAmazon
dc.description.sponsorshipMicrosoft
dc.description.sponsorshipGoogle
dc.description.sponsorshipTcL SYTRAL
dc.description.sponsorshipEuropean Language Resources Assoc
dc.description.sponsorshipOuaero
dc.description.sponsorshipImaginove
dc.description.sponsorshipVOCAPIA res
dc.description.sponsorshipAcapela
dc.description.sponsorshipSpeech ocean
dc.description.sponsorshipALDEBARAN
dc.description.sponsorshipOrange
dc.description.sponsorshipVecsys
dc.description.sponsorshipIBM Res
dc.description.sponsorshipRaytheon BBN Technol
dc.description.sponsorshipVoxygen
dc.identifier.citationHanilçi, C. vd. (2013). “Comparison of spectrum estimators in speaker verification: Mismatch conditions induced by vocal effort”. Interspeech, 14th Annual Conference of the International Speech Communication Association, 1-5, 2880-2884.
dc.identifier.endpage2884
dc.identifier.issn2308-457X
dc.identifier.issn978-1-62993-443-3
dc.identifier.scopus2-s2.0-84906242097
dc.identifier.startpage2880
dc.identifier.urihttp://hdl.handle.net/11452/32597
dc.identifier.volume1-5
dc.identifier.wos000395050001119
dc.indexed.wosCPCIS
dc.language.isoen
dc.publisherIsca-Int Speech Communication Assoc
dc.relation.collaborationYurt dışı
dc.relation.journalInterspeech, 14th Annual Conference of the International Speech Communication Association
dc.relation.publicationcategoryKonferans Öğesi - Uluslararası
dc.rightsinfo:eu-repo/semantics/closedAccess
dc.subjectComputer science
dc.subjectEngineering
dc.subjectSpeaker recognition
dc.subjectVocal effort mismatch
dc.subjectSpectrum estimation
dc.subjectLinear Prediction
dc.subjectModels
dc.subjectPoles
dc.subjectSpectrum analysis
dc.subjectFundamental frequencies
dc.subjectMel-frequency cepstral coefficients
dc.subjectRecognition accuracy
dc.subjectSpeaker recognition
dc.subjectSpeaker recognition evaluations
dc.subjectSpeaker verification
dc.subjectSpectrum estimation
dc.subjectVocal efforts
dc.subjectSpeech recognition
dc.subject.scopusSpeech Recognition; Language Recognition; Utterance
dc.subject.wosComputer science, artificial intelligence
dc.subject.wosEngineering, electrical & electronic
dc.titleComparison of spectrum estimators in speaker verification: Mismatch conditions induced by vocal effort
dc.typeProceedings Paper
dc.wos.quartileQ2
dspace.entity.typePublication
local.contributor.departmentMühendislik Fakültesi/Elektrik ve Elektronik Mühendisliği Bölümü
local.indexed.atScopus
local.indexed.atWOS

Files

License bundle

Now showing 1 - 1 of 1
Placeholder
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description: