Share on Facebook Tweet on Twitter Share on LinkedIn Share by email
Discriminative Learning in Speech Recognition

Xiaodong He and Li Deng

Abstract

In this paper, we study the objective functions of Maximum Mutual Information (MMI), Minimum Classification Error (MCE), and Minimum Phone/Word Error (MPE/MWE) for discriminative learning in speech recognition. We present an approach that unifies the objective functions of MMI, MCE and MPE/MWE in a common rational-function form. While the rational-function form of MMI has been known in the past, we provide a rigorous proof that the similar rational-function form exists for the objective functions of MCE and MPE/MWE. This allows the Growth Transformation (GT) or Extended Baum-Welch (EBW) based parameter optimization framework to be applied directly in discriminative learning. Prior to the current study, this framework was not directly applicable to MCE and MPE/MWE due to their lack of the appropriate rational-function form required by the GT/EBW-based parameter optimization method. In this paper, we include technical details on the derivation of the GT/EBW-based parameter optimization formulas for both discrete Hidden Markov Models (HMMs) and Continuous-Density HMMs (CDHMMs) in discriminative learning using MMI, MCE, and MPE/MWE criteria. For expository purposes, details on several related issues with practical significance are provided in Appendices.

Details

Publication typeTechReport
NumberMSR-TR-2007-129
Pages48
InstitutionMicrosoft Research
> Publications > Discriminative Learning in Speech Recognition