Binaural semi-blind dereverberation of noisy convoluted speech signals

Jong Hwan Lee; Sang Hoon Oh; Soo Young Lee

doi:10.1016/j.neucom.2008.07.005

Binaural semi-blind dereverberation of noisy convoluted speech signals

Jong Hwan Lee^*
, Sang Hoon Oh
, Soo Young Lee

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

6 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 language	English
Pages (from-to)	636-642
Number of pages	7
Journal	Neurocomputing
Volume	72
Issue number	1-3
DOIs	https://doi.org/10.1016/j.neucom.2008.07.005
Publication status	Published - 2008 Dec
Externally published	Yes

Bibliographical note

Funding Information:
The authors would like to thank the anonymous reviewers for their criticisms which improve this letter very much. Also, the authors thank Drs. Doh-Suk Kim and Hyung-Min Park for their fruitful discussions. This work was supported by the Brain Neuroinformatics Research Program sponsored by Korean Ministry of Commerce, Industry, and Energy.

Keywords

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

ASJC Scopus subject areas

Computer Science Applications
Cognitive Neuroscience
Artificial Intelligence

Access to Document

10.1016/j.neucom.2008.07.005

Cite this

@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 = "Lee, \{Jong Hwan\} and Oh, \{Sang Hoon\} and Lee, \{Soo Young\}",

note = "Funding Information: The authors would like to thank the anonymous reviewers for their criticisms which improve this letter very much. Also, the authors thank Drs. Doh-Suk Kim and Hyung-Min Park for their fruitful discussions. This work was supported by the Brain Neuroinformatics Research Program sponsored by Korean Ministry of Commerce, Industry, and Energy. ",

year = "2008",

month = dec,

doi = "10.1016/j.neucom.2008.07.005",

language = "English",

volume = "72",

pages = "636--642",

journal = "Neurocomputing",

issn = "0925-2312",

publisher = "Elsevier B.V.",

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

N1 - Funding Information: The authors would like to thank the anonymous reviewers for their criticisms which improve this letter very much. Also, the authors thank Drs. Doh-Suk Kim and Hyung-Min Park for their fruitful discussions. This work was supported by the Brain Neuroinformatics Research Program sponsored by Korean Ministry of Commerce, Industry, and Energy.

PY - 2008/12

Y1 - 2008/12

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 - https://www.scopus.com/pages/publications/55949137578

U2 - 10.1016/j.neucom.2008.07.005

DO - 10.1016/j.neucom.2008.07.005

M3 - Article

AN - SCOPUS:55949137578

SN - 0925-2312

VL - 72

SP - 636

EP - 642

JO - Neurocomputing

JF - Neurocomputing

IS - 1-3

ER -

Binaural semi-blind dereverberation of noisy convoluted speech signals

Abstract

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this