Robust text-independent speaker identification using Gaussian mixture speaker models | IEEE Journals & Magazine | IEEE Xplore