Please use this identifier to cite or link to this item: http://repo.lib.jfn.ac.lk/ujrr/handle/123456789/1899
Full metadata record
DC FieldValueLanguage
dc.contributor.authorAhilan, K.
dc.contributor.authorDean, D.
dc.contributor.authorSridharan, S.
dc.contributor.authorDominguez, J.G.
dc.contributor.authorRodriguez, J.G.
dc.contributor.authorRamos, D.
dc.date.accessioned2021-03-16T02:32:33Z
dc.date.accessioned2022-06-27T10:02:29Z-
dc.date.available2021-03-16T02:32:33Z
dc.date.available2022-06-27T10:02:29Z-
dc.date.issued2014
dc.identifier.citationImproving Short Utterance I-vector Speaker Verification using Utterance Variance Modelling and Compensation Techniquesen_US
dc.identifier.urihttp://repo.lib.jfn.ac.lk/ujrr/handle/123456789/1899-
dc.description.abstractThis paper proposes techniques to improve the performance of i-vector based speaker verification system when only short utterances are available. Short-length utterance i-vectors vary with speaker, session variations, and the phonetic content of the utterance. Well established methods such as linear discriminant analysis (LDA), source-normalized LDA (SN-LDA) and within-class covariance normalisation (WCCN) exist for compensating the session variation but we have identified the variability introduced by phonetic content due to utterance variation as an additional source of degradation when short-duration utterances are used. To compensate for utterance variations in short i-vector based speaker verification systems using cosine similarity scoring (CSS), we have introduced a short utterance variance normalization (SUVN) technique and a short utterance variance (SUV) modelling approach at the i-vector feature level. A combination of SUVN with LDA and SN-LDA is proposed to compensate the session and utterance variations and is shown to provide improvement in performance over the traditional approach of using LDA and/or SN-LDA followed by WCCN. An alternative approach is also introduced using the probabilistic linear discriminant analysis (PLDA) approach to directly model the SUV. The combination of SUVN, LDA and SN-LDA followed by SUV PLDA modelling provides an improvement over the baseline PLDA approach. We also show that for this combination of techniques, the utterance variation information needs to be artificially added to full-length i-vectors for PLDA modelling.en_US
dc.language.isoenen_US
dc.publisherElsevieren_US
dc.subjectSpeaker verificationen_US
dc.subjectI-vectoren_US
dc.titleImproving Short Utterance I-vector Speaker Verification using Utterance Variance Modelling and Compensation Techniquesen_US
dc.typeArticleen_US
Appears in Collections:Electrical & Electronic Engineering

Files in This Item:
File Description SizeFormat 
Improving Short Utterance I-vector Speaker Verification using.pdf128.85 kBAdobe PDFThumbnail
View/Open


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