Speaker identification using discrete wavelet packet transform technique with irregular decomposition

Jian-Da Wu, Bing Fu Lin

Research output: Contribution to journalArticle

54 Citations (Scopus)

Abstract

This paper presents the study of speaker identification for security systems based on the energy of speaker utterances. The proposed system consisted of a combination of signal pre-process, feature extraction using wavelet packet transform (WPT) and speaker identification using artificial neural network. In the signal pre-process, the amplitude of utterances, for a same sentence, were normalized for preventing an error estimation caused by speakers' change in volume. In the feature extraction, three conventional methods were considered in the experiments and compared with the irregular decomposition method in the proposed system. In order to verify the effect of the proposed system for identification, a general regressive neural network (GRNN) was used and compared in the experimental investigation. The experimental results demonstrated the effectiveness of the proposed speaker identification system and were compared with the discrete wavelet transform (DWT), conventional WPT and WPT in Mel scale.

Original languageEnglish
Pages (from-to)3136-3143
Number of pages8
JournalExpert Systems with Applications
Volume36
Issue number2 PART 2
DOIs
Publication statusPublished - 2009 Jan 1

Fingerprint

Feature extraction
Neural networks
Decomposition
Discrete wavelet transforms
Security systems
Error analysis
Identification (control systems)
Experiments

All Science Journal Classification (ASJC) codes

  • Engineering(all)
  • Computer Science Applications
  • Artificial Intelligence

Cite this

@article{73b844f84e234d8c8bd9460b6d14baa4,
title = "Speaker identification using discrete wavelet packet transform technique with irregular decomposition",
abstract = "This paper presents the study of speaker identification for security systems based on the energy of speaker utterances. The proposed system consisted of a combination of signal pre-process, feature extraction using wavelet packet transform (WPT) and speaker identification using artificial neural network. In the signal pre-process, the amplitude of utterances, for a same sentence, were normalized for preventing an error estimation caused by speakers' change in volume. In the feature extraction, three conventional methods were considered in the experiments and compared with the irregular decomposition method in the proposed system. In order to verify the effect of the proposed system for identification, a general regressive neural network (GRNN) was used and compared in the experimental investigation. The experimental results demonstrated the effectiveness of the proposed speaker identification system and were compared with the discrete wavelet transform (DWT), conventional WPT and WPT in Mel scale.",
author = "Jian-Da Wu and Lin, {Bing Fu}",
year = "2009",
month = "1",
day = "1",
doi = "10.1016/j.eswa.2008.01.038",
language = "English",
volume = "36",
pages = "3136--3143",
journal = "Expert Systems with Applications",
issn = "0957-4174",
publisher = "Elsevier Limited",
number = "2 PART 2",

}

Speaker identification using discrete wavelet packet transform technique with irregular decomposition. / Wu, Jian-Da; Lin, Bing Fu.

In: Expert Systems with Applications, Vol. 36, No. 2 PART 2, 01.01.2009, p. 3136-3143.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Speaker identification using discrete wavelet packet transform technique with irregular decomposition

AU - Wu, Jian-Da

AU - Lin, Bing Fu

PY - 2009/1/1

Y1 - 2009/1/1

N2 - This paper presents the study of speaker identification for security systems based on the energy of speaker utterances. The proposed system consisted of a combination of signal pre-process, feature extraction using wavelet packet transform (WPT) and speaker identification using artificial neural network. In the signal pre-process, the amplitude of utterances, for a same sentence, were normalized for preventing an error estimation caused by speakers' change in volume. In the feature extraction, three conventional methods were considered in the experiments and compared with the irregular decomposition method in the proposed system. In order to verify the effect of the proposed system for identification, a general regressive neural network (GRNN) was used and compared in the experimental investigation. The experimental results demonstrated the effectiveness of the proposed speaker identification system and were compared with the discrete wavelet transform (DWT), conventional WPT and WPT in Mel scale.

AB - This paper presents the study of speaker identification for security systems based on the energy of speaker utterances. The proposed system consisted of a combination of signal pre-process, feature extraction using wavelet packet transform (WPT) and speaker identification using artificial neural network. In the signal pre-process, the amplitude of utterances, for a same sentence, were normalized for preventing an error estimation caused by speakers' change in volume. In the feature extraction, three conventional methods were considered in the experiments and compared with the irregular decomposition method in the proposed system. In order to verify the effect of the proposed system for identification, a general regressive neural network (GRNN) was used and compared in the experimental investigation. The experimental results demonstrated the effectiveness of the proposed speaker identification system and were compared with the discrete wavelet transform (DWT), conventional WPT and WPT in Mel scale.

UR - http://www.scopus.com/inward/record.url?scp=56349117948&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=56349117948&partnerID=8YFLogxK

U2 - 10.1016/j.eswa.2008.01.038

DO - 10.1016/j.eswa.2008.01.038

M3 - Article

AN - SCOPUS:56349117948

VL - 36

SP - 3136

EP - 3143

JO - Expert Systems with Applications

JF - Expert Systems with Applications

SN - 0957-4174

IS - 2 PART 2

ER -