A comprehensive approach to blind source separation of speech mixtures

In this paper, we propose a novel algorithm for the separation of speech mixtures using two-microphone recordings, based on the combination of improved independent component analysis (ICA) and ideal binary mask (IBM). The improved FastICA algorithm reduce the number of Jacobian matrix, significantly reduce numbers of the convergence of the iteration. The basic signals are afterwards improved by the masks merging. In IBM, the separation stage is the iterative and repeated application until it reached the stopping criterion. The merging stage follow after that to reduce the probability of segregating the same signal repeatedly. The stereo property of the extracted speech signals can be maintained.