Speaker and Gender Normalization for Continuous-Density Hidden Markov Models

  • Alex Acero ,
  • Xuedong Huang

Proc. of the Int. Conf. on Acoustics, Speech, and Signal |

Published by IEEE

In this paper we describe a speaker-cluster normalization algorithm that we applied to both gender-normalization and speaker-normalization. To achieve parameter sharing the acoustic space is partitioned into classes. A maximum likelihood approach has been proposed under which the delta between the distribution mean and its corresponding acoustic class is mostly speaker-independent, whereas the means of the acoustic classes are mostly speaker-dependent. When applied to gender-normalization, the error rate reduction approaches that of a gender-dependent system but with half the number of parameters. For a speaker-normalized system, a 30% decrease in error rate was obtained in a batch recognition experiment in a context-dependent continuous-density HMM system.