Application of Empirical Mode Decomposition in Denoising a Speech Signal

Sonam Maheshwari, Dimpy

Abstract: This paper attempts to review and summarize the use of Empirical mode decomposition (EMD) for denoising a speech signal. EMD, introduced by Huang et al in 1998 gives time-frequency representation of non-linear and non-stationary signals. It decomposes a signal into a sum of finite zero-mean, oscillating components called as intrinsic mode functions based on the local time characteristics of the signal. Main essence of the method is its adaptive and data driven nature.

Keywords: Emirical mode decomposition, IMF, Wavelet transform, Hard and soft thresholding, Denoising