On the efficient speech feature extraction based on independent component analysis

Jong-Hwan Lee, Te Won Lee, Ho Young Jung, Soo Young Lee

Research output: Contribution to journalArticle

30 Citations (Scopus)

Abstract

A new efficient code for speech signals is proposed. To represent speech signals with minimum redundancy we use independent component analysis to adapt features (basis vectors) that efficiently encode the speech signals. The learned basis vectors are sparsely distributed and localized in both time and frequency. Time-frequency analysis of basis vectors shows the property similar with the critical bandwidth of human auditory system. Our results suggest that the obtained codes of speech signals are sparse and biologically plausible.

Original languageEnglish
Pages (from-to)235-245
Number of pages11
JournalNeural Processing Letters
Volume15
Issue number3
DOIs
Publication statusPublished - 2002 Jun 1
Externally publishedYes

Fingerprint

Independent component analysis
Feature extraction
Redundancy
Bandwidth

Keywords

  • Auditory system
  • Critical band
  • Feature extraction
  • Independent component analysis
  • Sparse code
  • Speech signal processing

ASJC Scopus subject areas

  • Artificial Intelligence
  • Neuroscience(all)

Cite this

On the efficient speech feature extraction based on independent component analysis. / Lee, Jong-Hwan; Lee, Te Won; Jung, Ho Young; Lee, Soo Young.

In: Neural Processing Letters, Vol. 15, No. 3, 01.06.2002, p. 235-245.

Research output: Contribution to journalArticle

Lee, Jong-Hwan ; Lee, Te Won ; Jung, Ho Young ; Lee, Soo Young. / On the efficient speech feature extraction based on independent component analysis. In: Neural Processing Letters. 2002 ; Vol. 15, No. 3. pp. 235-245.
@article{5def52753f30435195ee87b3ef97666f,
title = "On the efficient speech feature extraction based on independent component analysis",
abstract = "A new efficient code for speech signals is proposed. To represent speech signals with minimum redundancy we use independent component analysis to adapt features (basis vectors) that efficiently encode the speech signals. The learned basis vectors are sparsely distributed and localized in both time and frequency. Time-frequency analysis of basis vectors shows the property similar with the critical bandwidth of human auditory system. Our results suggest that the obtained codes of speech signals are sparse and biologically plausible.",
keywords = "Auditory system, Critical band, Feature extraction, Independent component analysis, Sparse code, Speech signal processing",
author = "Jong-Hwan Lee and Lee, {Te Won} and Jung, {Ho Young} and Lee, {Soo Young}",
year = "2002",
month = "6",
day = "1",
doi = "10.1023/A:1015777200976",
language = "English",
volume = "15",
pages = "235--245",
journal = "Neural Processing Letters",
issn = "1370-4621",
publisher = "Springer Netherlands",
number = "3",

}

TY - JOUR

T1 - On the efficient speech feature extraction based on independent component analysis

AU - Lee, Jong-Hwan

AU - Lee, Te Won

AU - Jung, Ho Young

AU - Lee, Soo Young

PY - 2002/6/1

Y1 - 2002/6/1

N2 - A new efficient code for speech signals is proposed. To represent speech signals with minimum redundancy we use independent component analysis to adapt features (basis vectors) that efficiently encode the speech signals. The learned basis vectors are sparsely distributed and localized in both time and frequency. Time-frequency analysis of basis vectors shows the property similar with the critical bandwidth of human auditory system. Our results suggest that the obtained codes of speech signals are sparse and biologically plausible.

AB - A new efficient code for speech signals is proposed. To represent speech signals with minimum redundancy we use independent component analysis to adapt features (basis vectors) that efficiently encode the speech signals. The learned basis vectors are sparsely distributed and localized in both time and frequency. Time-frequency analysis of basis vectors shows the property similar with the critical bandwidth of human auditory system. Our results suggest that the obtained codes of speech signals are sparse and biologically plausible.

KW - Auditory system

KW - Critical band

KW - Feature extraction

KW - Independent component analysis

KW - Sparse code

KW - Speech signal processing

UR - http://www.scopus.com/inward/record.url?scp=0036609363&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036609363&partnerID=8YFLogxK

U2 - 10.1023/A:1015777200976

DO - 10.1023/A:1015777200976

M3 - Article

AN - SCOPUS:0036609363

VL - 15

SP - 235

EP - 245

JO - Neural Processing Letters

JF - Neural Processing Letters

SN - 1370-4621

IS - 3

ER -