The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Une technique qui utilise un filtre d'erreur de prédiction linéaire (LPEF) et un filtre numérique adaptatif (ADF) pour obtenir une réduction du bruit dans une parole dégradée par un bruit de fond additif est proposée. On sait que les coefficients du LPEF convergent de telle sorte que le signal d'erreur de prédiction devient blanc. Etant donné qu'une parole voisée peut être représentée comme un signal périodique stationnaire sur un court intervalle de temps, la majeure partie de la parole voisée ne peut pas être incluse dans le signal d'erreur de prédiction du LPEF. En revanche, lorsque le signal d'entrée du LPEF est un bruit de fond, le signal d'erreur de prédiction devient blanc. En supposant que le bruit de fond est représenté comme généré par l'excitation d'un système linéaire avec un bruit blanc, nous pouvons alors reconstruire le bruit de fond à partir du signal d'erreur de prédiction en estimant la fonction de transfert du système de génération de bruit. Cette estimation est effectuée par l'ADF qui est utilisé comme identification du système. La réduction du bruit est obtenue en soustrayant le bruit reconstruit par l'ADF de la parole dégradée par le bruit de fond additif.
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.
Copier
Arata KAWAMURA, Kensaku FUJII, Yoshio ITOH, Yutaka FUKUI, "A New Noise Reduction Method Using Estimated Noise Spectrum" in IEICE TRANSACTIONS on Fundamentals,
vol. E85-A, no. 4, pp. 784-789, April 2002, doi: .
Abstract: A technique that uses a linear prediction error filter (LPEF) and an adaptive digital filter (ADF) to achieve noise reduction in a speech degraded by additive background noise is proposed. It is known that the coefficients of the LPEF converge such that the prediction error signal becomes white. Since a voiced speech can be represented as the stationary periodic signal over a short interval of time, most of voiced speech cannot be included in the prediction error signal of the LPEF. On the other hand, when the input signal of the LPEF is a background noise, the prediction error signal becomes white. Assuming that the background noise is represented as generate by exciting a linear system with a white noise, then we can reconstruct the background noise from the prediction error signal by estimating the transfer function of noise generation system. This estimation is performed by the ADF which is used as system identification. Noise reduction is achieved by subtracting the noise reconstructed by the ADF from the speech degraded by additive background noise.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/e85-a_4_784/_p
Copier
@ARTICLE{e85-a_4_784,
author={Arata KAWAMURA, Kensaku FUJII, Yoshio ITOH, Yutaka FUKUI, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={A New Noise Reduction Method Using Estimated Noise Spectrum},
year={2002},
volume={E85-A},
number={4},
pages={784-789},
abstract={A technique that uses a linear prediction error filter (LPEF) and an adaptive digital filter (ADF) to achieve noise reduction in a speech degraded by additive background noise is proposed. It is known that the coefficients of the LPEF converge such that the prediction error signal becomes white. Since a voiced speech can be represented as the stationary periodic signal over a short interval of time, most of voiced speech cannot be included in the prediction error signal of the LPEF. On the other hand, when the input signal of the LPEF is a background noise, the prediction error signal becomes white. Assuming that the background noise is represented as generate by exciting a linear system with a white noise, then we can reconstruct the background noise from the prediction error signal by estimating the transfer function of noise generation system. This estimation is performed by the ADF which is used as system identification. Noise reduction is achieved by subtracting the noise reconstructed by the ADF from the speech degraded by additive background noise.},
keywords={},
doi={},
ISSN={},
month={April},}
Copier
TY - JOUR
TI - A New Noise Reduction Method Using Estimated Noise Spectrum
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 784
EP - 789
AU - Arata KAWAMURA
AU - Kensaku FUJII
AU - Yoshio ITOH
AU - Yutaka FUKUI
PY - 2002
DO -
JO - IEICE TRANSACTIONS on Fundamentals
SN -
VL - E85-A
IS - 4
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - April 2002
AB - A technique that uses a linear prediction error filter (LPEF) and an adaptive digital filter (ADF) to achieve noise reduction in a speech degraded by additive background noise is proposed. It is known that the coefficients of the LPEF converge such that the prediction error signal becomes white. Since a voiced speech can be represented as the stationary periodic signal over a short interval of time, most of voiced speech cannot be included in the prediction error signal of the LPEF. On the other hand, when the input signal of the LPEF is a background noise, the prediction error signal becomes white. Assuming that the background noise is represented as generate by exciting a linear system with a white noise, then we can reconstruct the background noise from the prediction error signal by estimating the transfer function of noise generation system. This estimation is performed by the ADF which is used as system identification. Noise reduction is achieved by subtracting the noise reconstructed by the ADF from the speech degraded by additive background noise.
ER -