Single -Channel Audio Source Separation Using Adaptive EEMD and Local Margin Spectrum

Single -Channel Audio Source Separation Using Adaptive EEMD and Local Margin Spectrum

In this paper, a new and powerful method for blind audio source separation for single channel convolutive mixtures in a noisy environment is presented. This method is based onindependent sub-band analysis (ISA) in which Hilbert spectrum is employed. In the proposed algorithm, the adaptive EEMD is offeredto transfer the signal to the special intrinsic mode functions (IMF). We used the local margin spectrums (LoMS) as artificialobservations. Unlike using the EMD method, these observations of the proposed method do not contain any phantom sources. In order to make these independent observations, the Fast ICA method is used. A computer simulation is used to evaluate and compare the performance of the proposed method to the performances of twoother methods: the Hilbert–Huang EMD, and the Singular Spectrum Analysis (SSA) based methods. These comparisons arebased on Mutual orthogonality (MO) and Output SNR (OSNR) criteria. This simulation shows that the proposed algorithm improves the performance of the BSS system in a noisy environment.