Speaker Recognition Using Wavelet Cepstral Coefficient, I-Vector, and Cosine Distance Scoring and Its Application for Forensics

An important application of speaker recognition is forensics. However, the accuracy of speaker recognition in forensic cases often drops off rapidly because of the ill effect of ambient noise, variable channel, different duration of speech data, and so on. Therefore, finding a robust speaker recogni...

Full description

Saved in:
Bibliographic Details
Main Authors: Lei Lei, She Kun
Format: Article
Language:English
Published: Wiley 2016-01-01
Series:Journal of Electrical and Computer Engineering
Online Access:http://dx.doi.org/10.1155/2016/4908412
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850172447197757440
author Lei Lei
She Kun
author_facet Lei Lei
She Kun
author_sort Lei Lei
collection DOAJ
description An important application of speaker recognition is forensics. However, the accuracy of speaker recognition in forensic cases often drops off rapidly because of the ill effect of ambient noise, variable channel, different duration of speech data, and so on. Therefore, finding a robust speaker recognition model is very important for forensics. This paper builds a new speaker recognition model based on wavelet cepstral coefficient (WCC), i-vector, and cosine distance scoring (CDS). This model firstly uses the WCC to transform the speech into spectral feature vecors and then uses those spectral feature vectors to train the i-vectors that represent the speeches having different durations. CDS is used to compare the i-vectors to give out the evidence. Moreover, linear discriminant analysis (LDA) and the within-class covariance normalization (WCNN) are added to the CDS algorithm to deal with the channel variability problem. Finally, the likelihood ratio estimates the strength of the evidence. We use the TIMIT database to evaluate the performance of the proposed model. The experimental results show that the proposed model can effectively solve the troubles of forensic scenario, but the time cost of the method is high.
format Article
id doaj-art-c81d2e615b4b47c4b296aa9033469587
institution OA Journals
issn 2090-0147
2090-0155
language English
publishDate 2016-01-01
publisher Wiley
record_format Article
series Journal of Electrical and Computer Engineering
spelling doaj-art-c81d2e615b4b47c4b296aa90334695872025-08-20T02:20:06ZengWileyJournal of Electrical and Computer Engineering2090-01472090-01552016-01-01201610.1155/2016/49084124908412Speaker Recognition Using Wavelet Cepstral Coefficient, I-Vector, and Cosine Distance Scoring and Its Application for ForensicsLei Lei0She Kun1Laboratory of Cyberspace, School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, ChinaLaboratory of Cyberspace, School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu 610054, ChinaAn important application of speaker recognition is forensics. However, the accuracy of speaker recognition in forensic cases often drops off rapidly because of the ill effect of ambient noise, variable channel, different duration of speech data, and so on. Therefore, finding a robust speaker recognition model is very important for forensics. This paper builds a new speaker recognition model based on wavelet cepstral coefficient (WCC), i-vector, and cosine distance scoring (CDS). This model firstly uses the WCC to transform the speech into spectral feature vecors and then uses those spectral feature vectors to train the i-vectors that represent the speeches having different durations. CDS is used to compare the i-vectors to give out the evidence. Moreover, linear discriminant analysis (LDA) and the within-class covariance normalization (WCNN) are added to the CDS algorithm to deal with the channel variability problem. Finally, the likelihood ratio estimates the strength of the evidence. We use the TIMIT database to evaluate the performance of the proposed model. The experimental results show that the proposed model can effectively solve the troubles of forensic scenario, but the time cost of the method is high.http://dx.doi.org/10.1155/2016/4908412
spellingShingle Lei Lei
She Kun
Speaker Recognition Using Wavelet Cepstral Coefficient, I-Vector, and Cosine Distance Scoring and Its Application for Forensics
Journal of Electrical and Computer Engineering
title Speaker Recognition Using Wavelet Cepstral Coefficient, I-Vector, and Cosine Distance Scoring and Its Application for Forensics
title_full Speaker Recognition Using Wavelet Cepstral Coefficient, I-Vector, and Cosine Distance Scoring and Its Application for Forensics
title_fullStr Speaker Recognition Using Wavelet Cepstral Coefficient, I-Vector, and Cosine Distance Scoring and Its Application for Forensics
title_full_unstemmed Speaker Recognition Using Wavelet Cepstral Coefficient, I-Vector, and Cosine Distance Scoring and Its Application for Forensics
title_short Speaker Recognition Using Wavelet Cepstral Coefficient, I-Vector, and Cosine Distance Scoring and Its Application for Forensics
title_sort speaker recognition using wavelet cepstral coefficient i vector and cosine distance scoring and its application for forensics
url http://dx.doi.org/10.1155/2016/4908412
work_keys_str_mv AT leilei speakerrecognitionusingwaveletcepstralcoefficientivectorandcosinedistancescoringanditsapplicationforforensics
AT shekun speakerrecognitionusingwaveletcepstralcoefficientivectorandcosinedistancescoringanditsapplicationforforensics