Title :
Fast noise tracking based on recursive smoothing of MMSE noise power estimates
Author :
Erkelens, Jan S. ; Heusdens, Richard
Author_Institution :
Dept. of Mediamatics, Delft Univ. of Technol., Delft
fDate :
March 31 2008-April 4 2008
Abstract :
We consider estimation of the noise spectral variance from speech signals contaminated by highly nonstationary noise sources. In each time frame, for each frequency bin, the noise variance estimate is updated recursively with the minimum mean-square error (MMSE) estimate of the current noise power. For the estimation of the noise power, a spectral gain function is used, which is found by an iterative data-driven training method. The proposed noise tracking method can accurately track fast changes in noise level (up to about 10 dB/s). When compared to the minimum statistics method for various noise sources in a speech enhancement system, improvements in segmental signal-to-noise ratio of more than 1 dB are obtained.
Keywords :
acoustic noise; iterative methods; least mean squares methods; smoothing methods; spectral analysis; speech enhancement; tracking; MMSE; fast noise tracking; highly nonstationary noise sources; iterative data-driven training method; minimum mean-square error noise power estimation; noise spectral variance estimation; recursive smoothing; segmental signal-to-noise ratio; spectral gain function; speech enhancement system; speech signal; Acoustic noise; Frequency estimation; Iterative methods; Noise level; Noise reduction; Recursive estimation; Signal to noise ratio; Smoothing methods; Speech enhancement; Statistics; Speech enhancement; acoustic noise; least mean square methods; tracking filters;
Conference_Titel :
Acoustics, Speech and Signal Processing, 2008. ICASSP 2008. IEEE International Conference on
Conference_Location :
Las Vegas, NV
Print_ISBN :
978-1-4244-1483-3
Electronic_ISBN :
1520-6149
DOI :
10.1109/ICASSP.2008.4518749
