Top Message
Top Message
Back to Home Page  |  Settings   |  Sign In
Web Education
Viewing 1-1 of 1 total results
On-line Gaussian mixture modeling in the log-power domain ...
1. Introduction. The Gaussian mixture model (GMM) has been widely used in many applications of speech processing, including speech recognition, speaker verification and noise adaptation (The HTK Book, 1995, Reynolds and Rose, 1995, Acero, 1993, Burshtein and Gannot, 2002).In most applications, the GMM model is used to train the acoustical model or the speaker model or for noise compensation ...