Introduction - If you have any usage issues, please Google them yourself
(With reference to the article), can be separated and super-Gaussian linear mixed sub-Gaussian source signal to noise ratio based on the largest blind source separation algorithm, and can effectively separate the voice signal.