A New Noise Reduction Method Using Estimated Noise Spectrum

Kensaku FUJII
Yoshio ITOH
Yutaka FUKUI

IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences   Vol.E85-A    No.4    pp.784-789
Publication Date: 2002/04/01
Online ISSN: 
Print ISSN: 0916-8508
Type of Manuscript: Special Section PAPER (Special Section of Selected Papers from the 14th Workshop on Circuits and Systems in Karuizawa)
noise reduction,  linear prediction,  adaptive digital filter,  signal whitening,  noise reconstruction,  

Full Text: PDF>>
Buy this Article

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.