There are strong dependencies between wavelet coefficients of speech signal,in this article,based on that,a new corresponding nonlinear threshold function derived in Bayesian framework is proposed to decrease the effect of the ambient noise.Analysis of the data shows the effectiveness of the proposed method that it removes white noise more effectually and gets better edge preservation.
Performance Improvement of Adaptive Wavelet Thresholding for Speech Enhancement Using Generalized Gaussian Priors and Frame-Wise Context Modeling
