Speech Enhancement based on Empirical Mode Decomposition

P.N. Le, E. Ambikairajah, and V. Sethu (Australia)


Hilbert transform, speech enhancement, intrinsic mode function, empirical mode decomposition.


This paper presents a novel approach for single channel speech enhancement based on Empirical Mode Decomposition (EMD). Through EMD, noise and speech are distributed unequally in each intrinsic mode function (IMF). This enables us to weight the IMFs depending on their signal to noise ratios. Enhanced speech is obtained by combining the weighted IMFs. Experimental results indicate that the proposed method of noise reduction outperforms traditional speech enhancement methods.

Important Links:

Go Back