Empirical Mode Decomposition for Speech Enhancement

In this paper, we restrict in the use of the empirical mode decomposition (EMD) on the speech enhancement methods, the empirical mode decomposition (EMD) is based on an adaptive decomposition of the signal into oscillating components. A comparative study between the most recent EMD-based enhancement methods was performed. Experimental results show that the empirical mode decomposition can be used successfully in speech enhancement. The objective evaluation of the quality and intelligibility shows through an experimental evidence that the EMD-based speech enhancement methods provide a good performances.