Minimum classification error training for speaker identification using Gaussian mixture models based on multi-space probability distribution.

Chiyomi Miyajima, Keiichi Tokuda, Tadashi Kitamura

Browse the full Interspeech paper archive.