Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems

Chi Sang Jung, Moo Young Kim, Hong Goo Kang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

In this paper, an information theoretical approach to select features for speaker recognition systems is proposed. Conventional approaches having a fixed interval of analysis frames are not appropriate to represent dynamically varying characteristics of speech signals. To maximize the speakerrelated information varied by the characteristics of speech signals, we propose an information theory based feature selection method where features are selected to have minimum-redundancy with in selected features but maximumrelevancy to training speaker models. Experimental results verify that the proposed method reduces the error rates of speaker verification systems by 27.37 % in NIST 2002 database.

Original languageEnglish
Title of host publication2009 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings, ICASSP 2009
Pages4549-4552
Number of pages4
DOIs
Publication statusPublished - 2009 Sep 23
Event2009 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2009 - Taipei, Taiwan, Province of China
Duration: 2009 Apr 192009 Apr 24

Other

Other2009 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2009
CountryTaiwan, Province of China
CityTaipei
Period09/4/1909/4/24

Fingerprint

Redundancy
Feature extraction
Information theory

All Science Journal Classification (ASJC) codes

  • Software
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Jung, C. S., Kim, M. Y., & Kang, H. G. (2009). Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems. In 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings, ICASSP 2009 (pp. 4549-4552). [4960642] https://doi.org/10.1109/ICASSP.2009.4960642
Jung, Chi Sang ; Kim, Moo Young ; Kang, Hong Goo. / Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems. 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings, ICASSP 2009. 2009. pp. 4549-4552
@inproceedings{5fa574267ef3417bbe4c376c93da90cc,
title = "Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems",
abstract = "In this paper, an information theoretical approach to select features for speaker recognition systems is proposed. Conventional approaches having a fixed interval of analysis frames are not appropriate to represent dynamically varying characteristics of speech signals. To maximize the speakerrelated information varied by the characteristics of speech signals, we propose an information theory based feature selection method where features are selected to have minimum-redundancy with in selected features but maximumrelevancy to training speaker models. Experimental results verify that the proposed method reduces the error rates of speaker verification systems by 27.37 {\%} in NIST 2002 database.",
author = "Jung, {Chi Sang} and Kim, {Moo Young} and Kang, {Hong Goo}",
year = "2009",
month = "9",
day = "23",
doi = "10.1109/ICASSP.2009.4960642",
language = "English",
isbn = "9781424423545",
pages = "4549--4552",
booktitle = "2009 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings, ICASSP 2009",

}

Jung, CS, Kim, MY & Kang, HG 2009, Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems. in 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings, ICASSP 2009., 4960642, pp. 4549-4552, 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2009, Taipei, Taiwan, Province of China, 09/4/19. https://doi.org/10.1109/ICASSP.2009.4960642

Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems. / Jung, Chi Sang; Kim, Moo Young; Kang, Hong Goo.

2009 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings, ICASSP 2009. 2009. p. 4549-4552 4960642.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems

AU - Jung, Chi Sang

AU - Kim, Moo Young

AU - Kang, Hong Goo

PY - 2009/9/23

Y1 - 2009/9/23

N2 - In this paper, an information theoretical approach to select features for speaker recognition systems is proposed. Conventional approaches having a fixed interval of analysis frames are not appropriate to represent dynamically varying characteristics of speech signals. To maximize the speakerrelated information varied by the characteristics of speech signals, we propose an information theory based feature selection method where features are selected to have minimum-redundancy with in selected features but maximumrelevancy to training speaker models. Experimental results verify that the proposed method reduces the error rates of speaker verification systems by 27.37 % in NIST 2002 database.

AB - In this paper, an information theoretical approach to select features for speaker recognition systems is proposed. Conventional approaches having a fixed interval of analysis frames are not appropriate to represent dynamically varying characteristics of speech signals. To maximize the speakerrelated information varied by the characteristics of speech signals, we propose an information theory based feature selection method where features are selected to have minimum-redundancy with in selected features but maximumrelevancy to training speaker models. Experimental results verify that the proposed method reduces the error rates of speaker verification systems by 27.37 % in NIST 2002 database.

UR - http://www.scopus.com/inward/record.url?scp=70349213967&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70349213967&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2009.4960642

DO - 10.1109/ICASSP.2009.4960642

M3 - Conference contribution

AN - SCOPUS:70349213967

SN - 9781424423545

SP - 4549

EP - 4552

BT - 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings, ICASSP 2009

ER -

Jung CS, Kim MY, Kang HG. Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems. In 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings, ICASSP 2009. 2009. p. 4549-4552. 4960642 https://doi.org/10.1109/ICASSP.2009.4960642