Mel-Wiener Filter for Mel-LPC Based Speech Recognition
Summary :
This paper proposes a Mel-Wiener filter to enhance Mel-LPC spectra in the presence of additive noise. The transfer function of the proposed filter is defined by using a first-order all-pass filter instead of unit delay. The filter coefficients are estimated based on minimization of the sum of the square error on the linear frequency scale without applying the bilinear transformation and efficiently implemented in the autocorrelation domain. The proposed filter does not require any time-frequency conversion, which saves a large amount of computational load. The performance of the proposed system is comparable to that of ETSI AFE. The optimum filter order is found to be 3, and thus filtering is computationally inexpensive. The computational cost of the proposed system except VAD is 53% of ETSI AFE.
- Publication
- IEICE TRANSACTIONS on Information Vol.E90-D No.6 pp.935-942
- Publication Date
- 2007/06/01
- Publicized
- Online ISSN
- 1745-1361
- DOI
- 10.1093/ietisy/e90-d.6.935
- 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
Md. Babul ISLAM, Kazumasa YAMAMOTO, Hiroshi MATSUMOTO, "Mel-Wiener Filter for Mel-LPC Based Speech Recognition" in IEICE TRANSACTIONS on Information,
vol. E90-D, no. 6, pp. 935-942, June 2007, doi: 10.1093/ietisy/e90-d.6.935.
Abstract: This paper proposes a Mel-Wiener filter to enhance Mel-LPC spectra in the presence of additive noise. The transfer function of the proposed filter is defined by using a first-order all-pass filter instead of unit delay. The filter coefficients are estimated based on minimization of the sum of the square error on the linear frequency scale without applying the bilinear transformation and efficiently implemented in the autocorrelation domain. The proposed filter does not require any time-frequency conversion, which saves a large amount of computational load. The performance of the proposed system is comparable to that of ETSI AFE. The optimum filter order is found to be 3, and thus filtering is computationally inexpensive. The computational cost of the proposed system except VAD is 53% of ETSI AFE.
URL: https://global.ieice.org/en_transactions/information/10.1093/ietisy/e90-d.6.935/_p
Copy
@ARTICLE{e90-d_6_935,
author={Md. Babul ISLAM, Kazumasa YAMAMOTO, Hiroshi MATSUMOTO, },
journal={IEICE TRANSACTIONS on Information},
title={Mel-Wiener Filter for Mel-LPC Based Speech Recognition},
year={2007},
volume={E90-D},
number={6},
pages={935-942},
abstract={This paper proposes a Mel-Wiener filter to enhance Mel-LPC spectra in the presence of additive noise. The transfer function of the proposed filter is defined by using a first-order all-pass filter instead of unit delay. The filter coefficients are estimated based on minimization of the sum of the square error on the linear frequency scale without applying the bilinear transformation and efficiently implemented in the autocorrelation domain. The proposed filter does not require any time-frequency conversion, which saves a large amount of computational load. The performance of the proposed system is comparable to that of ETSI AFE. The optimum filter order is found to be 3, and thus filtering is computationally inexpensive. The computational cost of the proposed system except VAD is 53% of ETSI AFE.},
keywords={},
doi={10.1093/ietisy/e90-d.6.935},
ISSN={1745-1361},
month={June},}
Copy
TY - JOUR
TI - Mel-Wiener Filter for Mel-LPC Based Speech Recognition
T2 - IEICE TRANSACTIONS on Information
SP - 935
EP - 942
AU - Md. Babul ISLAM
AU - Kazumasa YAMAMOTO
AU - Hiroshi MATSUMOTO
PY - 2007
DO - 10.1093/ietisy/e90-d.6.935
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E90-D
IS - 6
JA - IEICE TRANSACTIONS on Information
Y1 - June 2007
AB - This paper proposes a Mel-Wiener filter to enhance Mel-LPC spectra in the presence of additive noise. The transfer function of the proposed filter is defined by using a first-order all-pass filter instead of unit delay. The filter coefficients are estimated based on minimization of the sum of the square error on the linear frequency scale without applying the bilinear transformation and efficiently implemented in the autocorrelation domain. The proposed filter does not require any time-frequency conversion, which saves a large amount of computational load. The performance of the proposed system is comparable to that of ETSI AFE. The optimum filter order is found to be 3, and thus filtering is computationally inexpensive. The computational cost of the proposed system except VAD is 53% of ETSI AFE.
ER -
English
