This site has been permanently archived. This is a static copy provided by the University of Southampton.

TY - GEN ID - cogprints6037 UR - http://cogprints.org/6037/ A1 - Kanevsky, Dr Dimitri A1 - Povey, Dr Daniel A1 - Ramabhadran, Dr Bhuvana A1 - Rish, Dr Irina A1 - Sainath, Dr Tara TI - A New Family of Extended Baum-Welch Update Rules Y1 - 2008/04/27/ N2 - In this paper, we consider a generalization of the state-of-art discriminative method for optimizing the conditional likelihood in Hidden Markov Models (HMMs), called the Extended Baum-Welch (EBW) algorithm, that has had significant impact on the speech recognition community. We propose a generalized form of EBW update rules that can be associated with a weighted sum of updated and initial models, and demonstrate that using novel update rules can significantly speed up parameter estimation for Gaussian mixtures. AV - public KW - Extended Baum Welch KW - optimization KW - speech ER -