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
Propusemos um método de redução de ruído baseado em um sistema de reconstrução de ruído (NRS). O NRS usa um filtro de erro de predição linear (LPEF) e um filtro de reconstrução de ruído (NRF) que estima o ruído de fundo por identificação do sistema. No caso de ser utilizado um tamanho de passo fixo para actualizar os coeficientes de derivação do NRF, é difícil reduzir o ruído de fundo enquanto se mantém a elevada qualidade da fala melhorada. Para resolver o problema, é proposto um tamanho de passo variável. Ele faz uso da correlação cruzada entre um sinal de entrada e um sinal de fala aprimorado. Em uma seção de fala, um tamanho de passo variável torna-se pequeno para não estimar a fala, por outro lado, grande para rastrear o ruído de fundo em uma seção sem fala.
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.
Copiar
Naoto SASAOKA, Masatoshi WATANABE, Yoshio ITOH, Kensaku FUJII, "A Variable Step Size Algorithm for Speech Noise Reduction Method Based on Noise Reconstruction System" in IEICE TRANSACTIONS on Fundamentals,
vol. E92-A, no. 1, pp. 244-251, January 2009, doi: 10.1587/transfun.E92.A.244.
Abstract: We have proposed a noise reduction method based on a noise reconstruction system (NRS). The NRS uses a linear prediction error filter (LPEF) and a noise reconstruction filter (NRF) which estimates background noise by system identification. In case a fixed step size for updating tap coefficients of the NRF is used, it is difficult to reduce background noise while maintaining the high quality of enhanced speech. In order to solve the problem, a variable step size is proposed. It makes use of cross-correlation between an input signal and an enhanced speech signal. In a speech section, a variable step size becomes small so as not to estimate speech, on the other hand, large to track the background noise in a non-speech section.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.E92.A.244/_p
Copiar
@ARTICLE{e92-a_1_244,
author={Naoto SASAOKA, Masatoshi WATANABE, Yoshio ITOH, Kensaku FUJII, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={A Variable Step Size Algorithm for Speech Noise Reduction Method Based on Noise Reconstruction System},
year={2009},
volume={E92-A},
number={1},
pages={244-251},
abstract={We have proposed a noise reduction method based on a noise reconstruction system (NRS). The NRS uses a linear prediction error filter (LPEF) and a noise reconstruction filter (NRF) which estimates background noise by system identification. In case a fixed step size for updating tap coefficients of the NRF is used, it is difficult to reduce background noise while maintaining the high quality of enhanced speech. In order to solve the problem, a variable step size is proposed. It makes use of cross-correlation between an input signal and an enhanced speech signal. In a speech section, a variable step size becomes small so as not to estimate speech, on the other hand, large to track the background noise in a non-speech section.},
keywords={},
doi={10.1587/transfun.E92.A.244},
ISSN={1745-1337},
month={January},}
Copiar
TY - JOUR
TI - A Variable Step Size Algorithm for Speech Noise Reduction Method Based on Noise Reconstruction System
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 244
EP - 251
AU - Naoto SASAOKA
AU - Masatoshi WATANABE
AU - Yoshio ITOH
AU - Kensaku FUJII
PY - 2009
DO - 10.1587/transfun.E92.A.244
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E92-A
IS - 1
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - January 2009
AB - We have proposed a noise reduction method based on a noise reconstruction system (NRS). The NRS uses a linear prediction error filter (LPEF) and a noise reconstruction filter (NRF) which estimates background noise by system identification. In case a fixed step size for updating tap coefficients of the NRF is used, it is difficult to reduce background noise while maintaining the high quality of enhanced speech. In order to solve the problem, a variable step size is proposed. It makes use of cross-correlation between an input signal and an enhanced speech signal. In a speech section, a variable step size becomes small so as not to estimate speech, on the other hand, large to track the background noise in a non-speech section.
ER -