An algorithm for blind source separation based on wavelet probability density estimation
Ying Gao · Journal of Guangzhou University · 2006
In this paper,an algorithm for linear blind source separation is presented by applying wavelet probability density estimation.Instead of using nonlinear functions,the proposed algorithm use wavelet probability density estimation to estimate the score functions of the signals directly.The proposed algorithm has ability to separate hybrid mixtures that contain both super Gaussian and sub Gaussian sources,and also a simple implementation.Computer simulation results show that the proposed algorithm has good performance.