Isolated word recognition in reverberant environments

Shu Guang Wang; Xiang Yang Zeng; Qiang Wang

doi:10.1109/ICSPCC.2011.6061575

Isolated word recognition in reverberant environments

Shu Guang Wang, Xiang Yang Zeng, Qiang Wang

School of Marine Science and Technology

Northwestern Polytechnical University Xian

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

The additive noise and channel distortion caused by reverberation can degrade the performance of isolated word recognition(IWR), and have become the key constraint to the applications of IWR. In this paper, we present a reverberation robust isolated word recognition method. By using the relative autocorrelation sequences (RAS) based voice activity detection, influences of additive noise can be eliminated. To reduce the channel distortion Cepstral mean subtraction (CMS) is employed in Mel frequency cepstral coefficients (MFCC) extraction. And Gaussian mixture model (GMM) is used for the statistical modeling. The performance of the presented method in various reverberation conditions was evaluated by the recognition experiments.

Original language	English
Title of host publication	2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011
DOIs	https://doi.org/10.1109/ICSPCC.2011.6061575
State	Published - 2011
Event	2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011 - Xi'an, China Duration: 14 Sep 2011 → 16 Sep 2011

Publication series

Name	2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011

Conference

Conference	2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011
Country/Territory	China
City	Xi'an
Period	14/09/11 → 16/09/11

Keywords

cepstral mean subtraction
Gaussian mixture model
isolated word recognition
Mel frequency cepstral coefficients
relative autocorrelation sequences
reverberation

Access to Document

10.1109/ICSPCC.2011.6061575

Cite this

@inproceedings{37f210592f164eceab15f5270887f5f7,

title = "Isolated word recognition in reverberant environments",

abstract = "The additive noise and channel distortion caused by reverberation can degrade the performance of isolated word recognition(IWR), and have become the key constraint to the applications of IWR. In this paper, we present a reverberation robust isolated word recognition method. By using the relative autocorrelation sequences (RAS) based voice activity detection, influences of additive noise can be eliminated. To reduce the channel distortion Cepstral mean subtraction (CMS) is employed in Mel frequency cepstral coefficients (MFCC) extraction. And Gaussian mixture model (GMM) is used for the statistical modeling. The performance of the presented method in various reverberation conditions was evaluated by the recognition experiments.",

keywords = "cepstral mean subtraction, Gaussian mixture model, isolated word recognition, Mel frequency cepstral coefficients, relative autocorrelation sequences, reverberation",

author = "Wang, {Shu Guang} and Zeng, {Xiang Yang} and Qiang Wang",

year = "2011",

doi = "10.1109/ICSPCC.2011.6061575",

language = "英语",

isbn = "9781457708947",

series = "2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011",

booktitle = "2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011",

note = "2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011 ; Conference date: 14-09-2011 Through 16-09-2011",

}

Wang, SG, Zeng, XY & Wang, Q 2011, Isolated word recognition in reverberant environments. in 2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011., 6061575, 2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011, 2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011, Xi'an, China, 14/09/11. https://doi.org/10.1109/ICSPCC.2011.6061575

Isolated word recognition in reverberant environments. / Wang, Shu Guang; Zeng, Xiang Yang; Wang, Qiang.
2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011. 2011. 6061575 (2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Isolated word recognition in reverberant environments

AU - Wang, Shu Guang

AU - Zeng, Xiang Yang

AU - Wang, Qiang

PY - 2011

Y1 - 2011

N2 - The additive noise and channel distortion caused by reverberation can degrade the performance of isolated word recognition(IWR), and have become the key constraint to the applications of IWR. In this paper, we present a reverberation robust isolated word recognition method. By using the relative autocorrelation sequences (RAS) based voice activity detection, influences of additive noise can be eliminated. To reduce the channel distortion Cepstral mean subtraction (CMS) is employed in Mel frequency cepstral coefficients (MFCC) extraction. And Gaussian mixture model (GMM) is used for the statistical modeling. The performance of the presented method in various reverberation conditions was evaluated by the recognition experiments.

AB - The additive noise and channel distortion caused by reverberation can degrade the performance of isolated word recognition(IWR), and have become the key constraint to the applications of IWR. In this paper, we present a reverberation robust isolated word recognition method. By using the relative autocorrelation sequences (RAS) based voice activity detection, influences of additive noise can be eliminated. To reduce the channel distortion Cepstral mean subtraction (CMS) is employed in Mel frequency cepstral coefficients (MFCC) extraction. And Gaussian mixture model (GMM) is used for the statistical modeling. The performance of the presented method in various reverberation conditions was evaluated by the recognition experiments.

KW - cepstral mean subtraction

KW - Gaussian mixture model

KW - isolated word recognition

KW - Mel frequency cepstral coefficients

KW - relative autocorrelation sequences

KW - reverberation

UR - http://www.scopus.com/inward/record.url?scp=81055138864&partnerID=8YFLogxK

U2 - 10.1109/ICSPCC.2011.6061575

DO - 10.1109/ICSPCC.2011.6061575

M3 - 会议稿件

AN - SCOPUS:81055138864

SN - 9781457708947

T3 - 2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011

BT - 2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011

T2 - 2011 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2011

Y2 - 14 September 2011 through 16 September 2011

ER -

Isolated word recognition in reverberant environments

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this