Binaural semi-blind dereverberation of noisy convoluted speech signals

Jong-Hwan Lee, Sang Hoon Oh, Soo Young Lee

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

In order to overcome a limited performance of a conventional monaural model, this letter proposes a binaural blind dereverberation model. Its learning rule is derived using a blind least-squares measure by exploiting higher-order characteristics of output components. In order to prevent an unwanted whitening of speech signal, we adopt a semi-blind approach by employing a pre-determined whitening filter. The proposed model is evaluated using several simulated conditions and the results show better speech quality than those of the monaural model. The applicability of the model to the real environment is also shown by applying to real-recorded data. Especially, the proposed model attains much improved word error rates from 13.9 ± 5.7 (%) to 4.1 ± 3.5 (%) across 13 speakers for testing in the real speech recognition experiments.

Original languageEnglish
Pages (from-to)636-642
Number of pages7
JournalNeurocomputing
Volume72
Issue number1-3
DOIs
Publication statusPublished - 2008 Dec 1
Externally publishedYes

Fingerprint

Least-Squares Analysis
Learning
Speech recognition
Testing
Experiments
Recognition (Psychology)

Keywords

  • Automatic speech recognition
  • Blind deconvolution
  • Blind dereverberation
  • Blind least squares
  • Independent component analysis
  • Speech enhancement

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Cognitive Neuroscience

Cite this

Binaural semi-blind dereverberation of noisy convoluted speech signals. / Lee, Jong-Hwan; Oh, Sang Hoon; Lee, Soo Young.

In: Neurocomputing, Vol. 72, No. 1-3, 01.12.2008, p. 636-642.

Research output: Contribution to journalArticle

Lee, Jong-Hwan ; Oh, Sang Hoon ; Lee, Soo Young. / Binaural semi-blind dereverberation of noisy convoluted speech signals. In: Neurocomputing. 2008 ; Vol. 72, No. 1-3. pp. 636-642.
@article{54e62aa50cfa44b88e269264b901c3d8,
title = "Binaural semi-blind dereverberation of noisy convoluted speech signals",
abstract = "In order to overcome a limited performance of a conventional monaural model, this letter proposes a binaural blind dereverberation model. Its learning rule is derived using a blind least-squares measure by exploiting higher-order characteristics of output components. In order to prevent an unwanted whitening of speech signal, we adopt a semi-blind approach by employing a pre-determined whitening filter. The proposed model is evaluated using several simulated conditions and the results show better speech quality than those of the monaural model. The applicability of the model to the real environment is also shown by applying to real-recorded data. Especially, the proposed model attains much improved word error rates from 13.9 ± 5.7 ({\%}) to 4.1 ± 3.5 ({\%}) across 13 speakers for testing in the real speech recognition experiments.",
keywords = "Automatic speech recognition, Blind deconvolution, Blind dereverberation, Blind least squares, Independent component analysis, Speech enhancement",
author = "Jong-Hwan Lee and Oh, {Sang Hoon} and Lee, {Soo Young}",
year = "2008",
month = "12",
day = "1",
doi = "10.1016/j.neucom.2008.07.005",
language = "English",
volume = "72",
pages = "636--642",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",
number = "1-3",

}

TY - JOUR

T1 - Binaural semi-blind dereverberation of noisy convoluted speech signals

AU - Lee, Jong-Hwan

AU - Oh, Sang Hoon

AU - Lee, Soo Young

PY - 2008/12/1

Y1 - 2008/12/1

N2 - In order to overcome a limited performance of a conventional monaural model, this letter proposes a binaural blind dereverberation model. Its learning rule is derived using a blind least-squares measure by exploiting higher-order characteristics of output components. In order to prevent an unwanted whitening of speech signal, we adopt a semi-blind approach by employing a pre-determined whitening filter. The proposed model is evaluated using several simulated conditions and the results show better speech quality than those of the monaural model. The applicability of the model to the real environment is also shown by applying to real-recorded data. Especially, the proposed model attains much improved word error rates from 13.9 ± 5.7 (%) to 4.1 ± 3.5 (%) across 13 speakers for testing in the real speech recognition experiments.

AB - In order to overcome a limited performance of a conventional monaural model, this letter proposes a binaural blind dereverberation model. Its learning rule is derived using a blind least-squares measure by exploiting higher-order characteristics of output components. In order to prevent an unwanted whitening of speech signal, we adopt a semi-blind approach by employing a pre-determined whitening filter. The proposed model is evaluated using several simulated conditions and the results show better speech quality than those of the monaural model. The applicability of the model to the real environment is also shown by applying to real-recorded data. Especially, the proposed model attains much improved word error rates from 13.9 ± 5.7 (%) to 4.1 ± 3.5 (%) across 13 speakers for testing in the real speech recognition experiments.

KW - Automatic speech recognition

KW - Blind deconvolution

KW - Blind dereverberation

KW - Blind least squares

KW - Independent component analysis

KW - Speech enhancement

UR - http://www.scopus.com/inward/record.url?scp=55949137578&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=55949137578&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2008.07.005

DO - 10.1016/j.neucom.2008.07.005

M3 - Article

AN - SCOPUS:55949137578

VL - 72

SP - 636

EP - 642

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - 1-3

ER -