IEICE TRANSACTIONS on Information
Cepstral Statistics Compensation and Normalization Using Online Pseudo Stereo Codebooks for Robust Speech Recognition in Additive Noise Environments
Summary :
This paper proposes several cepstral statistics compensation and normalization algorithms which alleviate the effect of additive noise on cepstral features for speech recognition. The algorithms are simple yet efficient noise reduction techniques that use online-constructed pseudo-stereo codebooks to evaluate the statistics in both clean and noisy environments. The process yields transformations for both clean speech cepstra and noise-corrupted speech cepstra, or for noise-corrupted speech cepstra only, so that the statistics of the transformed speech cepstra are similar for both environments. Experimental results show that these codebook-based algorithms can provide significant performance gains compared to results obtained by using conventional utterance-based normalization approaches. The proposed codebook-based cesptral mean and variance normalization (C-CMVN), linear least squares (LLS) and quadratic least squares (QLS) outperform utterance-based CMVN (U-CMVN) by 26.03%, 22.72% and 27.48%, respectively, in relative word error rate reduction for experiments conducted on Test Set A of the Aurora-2 digit database.
- Publication
- IEICE TRANSACTIONS on Information Vol.E91-D No.2 pp.296-311
- Publication Date
- 2008/02/01
- Publicized
- Online ISSN
- 1745-1361
- DOI
- 10.1093/ietisy/e91-d.2.296
- Type of Manuscript
- PAPER
- Category
- Speech and Hearing
Authors
Keyword
Latest Issue
Copyrights notice of machine-translated contents
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Cite this
Copy
Jeih-weih HUNG, "Cepstral Statistics Compensation and Normalization Using Online Pseudo Stereo Codebooks for Robust Speech Recognition in Additive Noise Environments" in IEICE TRANSACTIONS on Information,
vol. E91-D, no. 2, pp. 296-311, February 2008, doi: 10.1093/ietisy/e91-d.2.296.
Abstract: This paper proposes several cepstral statistics compensation and normalization algorithms which alleviate the effect of additive noise on cepstral features for speech recognition. The algorithms are simple yet efficient noise reduction techniques that use online-constructed pseudo-stereo codebooks to evaluate the statistics in both clean and noisy environments. The process yields transformations for both clean speech cepstra and noise-corrupted speech cepstra, or for noise-corrupted speech cepstra only, so that the statistics of the transformed speech cepstra are similar for both environments. Experimental results show that these codebook-based algorithms can provide significant performance gains compared to results obtained by using conventional utterance-based normalization approaches. The proposed codebook-based cesptral mean and variance normalization (C-CMVN), linear least squares (LLS) and quadratic least squares (QLS) outperform utterance-based CMVN (U-CMVN) by 26.03%, 22.72% and 27.48%, respectively, in relative word error rate reduction for experiments conducted on Test Set A of the Aurora-2 digit database.
URL: https://global.ieice.org/en_transactions/information/10.1093/ietisy/e91-d.2.296/_p
Copy
@ARTICLE{e91-d_2_296,
author={Jeih-weih HUNG, },
journal={IEICE TRANSACTIONS on Information},
title={Cepstral Statistics Compensation and Normalization Using Online Pseudo Stereo Codebooks for Robust Speech Recognition in Additive Noise Environments},
year={2008},
volume={E91-D},
number={2},
pages={296-311},
abstract={This paper proposes several cepstral statistics compensation and normalization algorithms which alleviate the effect of additive noise on cepstral features for speech recognition. The algorithms are simple yet efficient noise reduction techniques that use online-constructed pseudo-stereo codebooks to evaluate the statistics in both clean and noisy environments. The process yields transformations for both clean speech cepstra and noise-corrupted speech cepstra, or for noise-corrupted speech cepstra only, so that the statistics of the transformed speech cepstra are similar for both environments. Experimental results show that these codebook-based algorithms can provide significant performance gains compared to results obtained by using conventional utterance-based normalization approaches. The proposed codebook-based cesptral mean and variance normalization (C-CMVN), linear least squares (LLS) and quadratic least squares (QLS) outperform utterance-based CMVN (U-CMVN) by 26.03%, 22.72% and 27.48%, respectively, in relative word error rate reduction for experiments conducted on Test Set A of the Aurora-2 digit database.},
keywords={},
doi={10.1093/ietisy/e91-d.2.296},
ISSN={1745-1361},
month={February},}
Copy
TY - JOUR
TI - Cepstral Statistics Compensation and Normalization Using Online Pseudo Stereo Codebooks for Robust Speech Recognition in Additive Noise Environments
T2 - IEICE TRANSACTIONS on Information
SP - 296
EP - 311
AU - Jeih-weih HUNG
PY - 2008
DO - 10.1093/ietisy/e91-d.2.296
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E91-D
IS - 2
JA - IEICE TRANSACTIONS on Information
Y1 - February 2008
AB - This paper proposes several cepstral statistics compensation and normalization algorithms which alleviate the effect of additive noise on cepstral features for speech recognition. The algorithms are simple yet efficient noise reduction techniques that use online-constructed pseudo-stereo codebooks to evaluate the statistics in both clean and noisy environments. The process yields transformations for both clean speech cepstra and noise-corrupted speech cepstra, or for noise-corrupted speech cepstra only, so that the statistics of the transformed speech cepstra are similar for both environments. Experimental results show that these codebook-based algorithms can provide significant performance gains compared to results obtained by using conventional utterance-based normalization approaches. The proposed codebook-based cesptral mean and variance normalization (C-CMVN), linear least squares (LLS) and quadratic least squares (QLS) outperform utterance-based CMVN (U-CMVN) by 26.03%, 22.72% and 27.48%, respectively, in relative word error rate reduction for experiments conducted on Test Set A of the Aurora-2 digit database.
ER -
English
