Speech Enhancement Based on MAP Estimation Using a Variable Speech Distribution

Yuta TSUKAMOTO; Arata KAWAMURA; Youji IIGUNI

doi:10.1093/ietfec/e90-a.8.1587

Speech Enhancement Based on MAP Estimation Using a Variable Speech Distribution

Yuta TSUKAMOTO, Arata KAWAMURA, Youji IIGUNI

Full Text Views

0

Share
Cite this

Summary :

In this paper, a novel speech enhancement algorithm based on the MAP estimation is proposed. The proposed speech enhancer adaptively changes the speech spectral density used in the MAP estimation according to the sum of the observed power spectra. In a speech segment, the speech spectral density approaches to Rayleigh distribution to keep the quality of the enhanced speech. While in a non-speech segment, it approaches to an exponential distribution to reduce noise effectively. Furthermore, when the noise is super-Gaussian, we modify the width of Gaussian so that the Gaussian model with the modified width approximates the distribution of the super-Gaussian noise. This technique is effective in suppressing residual noise well. From computer experiments, we confirm the effectiveness of the proposed method.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E90-A No.8 pp.1587-1593

Publication Date: 2007/08/01

Publicized

Online ISSN: 1745-1337

DOI: 10.1093/ietfec/e90-a.8.1587

Type of Manuscript: Special Section PAPER (Special Section on Papers Selected from the 21st Symposium on Signal Processing)

Category

Cite this

Copy

Yuta TSUKAMOTO, Arata KAWAMURA, Youji IIGUNI, "Speech Enhancement Based on MAP Estimation Using a Variable Speech Distribution" in IEICE TRANSACTIONS on Fundamentals, vol. E90-A, no. 8, pp. 1587-1593, August 2007, doi: 10.1093/ietfec/e90-a.8.1587.
Abstract: In this paper, a novel speech enhancement algorithm based on the MAP estimation is proposed. The proposed speech enhancer adaptively changes the speech spectral density used in the MAP estimation according to the sum of the observed power spectra. In a speech segment, the speech spectral density approaches to Rayleigh distribution to keep the quality of the enhanced speech. While in a non-speech segment, it approaches to an exponential distribution to reduce noise effectively. Furthermore, when the noise is super-Gaussian, we modify the width of Gaussian so that the Gaussian model with the modified width approximates the distribution of the super-Gaussian noise. This technique is effective in suppressing residual noise well. From computer experiments, we confirm the effectiveness of the proposed method.
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1093/ietfec/e90-a.8.1587/_p

Copy

@ARTICLE{e90-a_8_1587,
author={Yuta TSUKAMOTO, Arata KAWAMURA, Youji IIGUNI, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Speech Enhancement Based on MAP Estimation Using a Variable Speech Distribution},
year={2007},
volume={E90-A},
number={8},
pages={1587-1593},
abstract={In this paper, a novel speech enhancement algorithm based on the MAP estimation is proposed. The proposed speech enhancer adaptively changes the speech spectral density used in the MAP estimation according to the sum of the observed power spectra. In a speech segment, the speech spectral density approaches to Rayleigh distribution to keep the quality of the enhanced speech. While in a non-speech segment, it approaches to an exponential distribution to reduce noise effectively. Furthermore, when the noise is super-Gaussian, we modify the width of Gaussian so that the Gaussian model with the modified width approximates the distribution of the super-Gaussian noise. This technique is effective in suppressing residual noise well. From computer experiments, we confirm the effectiveness of the proposed method.},
keywords={},
doi={10.1093/ietfec/e90-a.8.1587},
ISSN={1745-1337},
month={August},}

Copy

TY - JOUR
TI - Speech Enhancement Based on MAP Estimation Using a Variable Speech Distribution
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1587
EP - 1593
AU - Yuta TSUKAMOTO
AU - Arata KAWAMURA
AU - Youji IIGUNI
PY - 2007
DO - 10.1093/ietfec/e90-a.8.1587
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E90-A
IS - 8
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - August 2007
AB - In this paper, a novel speech enhancement algorithm based on the MAP estimation is proposed. The proposed speech enhancer adaptively changes the speech spectral density used in the MAP estimation according to the sum of the observed power spectra. In a speech segment, the speech spectral density approaches to Rayleigh distribution to keep the quality of the enhanced speech. While in a non-speech segment, it approaches to an exponential distribution to reduce noise effectively. Furthermore, when the noise is super-Gaussian, we modify the width of Gaussian so that the Gaussian model with the modified width approximates the distribution of the super-Gaussian noise. This technique is effective in suppressing residual noise well. From computer experiments, we confirm the effectiveness of the proposed method.
ER -