Speech Denoising by Adaptive Filter LMS in the EMD Framework
Haifa Touati, Kais Khaldi · 2018
This work introduces a new speech de noising method, which combines adaptive Least Mean Squares (LMS) filter with Empirical Mode Decomposition (EMD). The basic idea of this method is the signal reconstruction with IMFs previously filtered, using the adaptive (LMS) filter. The LMS algorithm has been extensively used in many applications as a consequence of its simplicity and robustness. Thanks to the data driven decomposition of the EMD, the application of the adaptive filter on the IMFs gives better results than use alone LMS filtering of the noisy signal. The proposed EMD-LMS denoising method is applied to noisy speech signal with different noise levels and the results are compared with those of the LMS filter. The obtained results show that the proposed denoising schemes perform better than the LMS filter. The study is limited to signals corrupted by additive white Gaussian noise.