GMM modeling of speech signal for speech enhancement

Hi fnds,


I wanted to find the Gaussian Mixture Model (GMM) of speech frames (i.e, row vector of 960 samples) in matlab to implement one speech enhancement paper.I tried some statistical procedure that i got in internet and also tried the matlab instructions "gmdistribtuion" and "gmdistribution.fit",But am getting errors in either the ways.So if any of u have idea about this Gaussian Mixture Model(GMM), plz help me out.