I have problem with GMM testing. I want to know how we can compute the probability of testing vectors by trained GMMs? Let's ask in other way: do we have to compute the means and covariances of GMMS and then, based on them, apply Gaussian PDF for computing the test probabilities? How can we calculate the universal background model (UBM) using MATLAB? I will appreciate, if you put the MATLAB code for better understanding.

More Mohammad Ali Nematollahi's questions See All
Similar questions and discussions