Performance of Discrete Wavelet Transform (DWT) Based Speech Denoising in Impulsive and Gaussian Noise

Abstract

The aim of this paper is to investigate the effect of impulsive noise on the performance of the speech denoising using Discrete Wavelet Transform (DWT). The employed model of impulsive noise consists of Bernoulli distributed impulse arrivals and Gaussian distributed amplitudes of the impulses. In this study DWT algorithm has been applied for the suppression of ambient noise. This method is based on thresholding the wavelet coefficients, that can be done by standard deviation method for each frame by level dependent thresholding using different types of threshold (semisoft, hard soft and super soft) in channel contains impulsive and Gaussian noise together. The results of simulation indicate that using discrete wavelet transform in speech denoising application provides a good quality and semisoft threshold gives the best performance.

Keywords

• .