A novel online algorithm for blind source separation with unknown number of sources
Yanli Li, Bo Li · 2011
A novel algorithm which can online estimate number of sources and separate sources is proposed to solve the Blind Source Separation problem with unknown number of sources. First create a discriminator and threshold, and adjust the estimating number of sources via comparing the size of discriminator and threshold gradually to converge at the true number. Then use a modified natural gradient algorithm to separate the sources. Computer simulation verify the algorithm is effective and perform better than traditional algorithm.