Text-dependent speaker identification based on input/output HMMs: An empirical study

摘要

In this paper, we explore theInput/Output HMM (IOHMM) architecture for a substantial problem, that of text-dependent speaker identification. For subnetworks modeled with generalized linear models, we extend the IRLS algorithm to the M-step of the corresponding EM algorithm. Experimental results show that the improved EM algorithm yields significantly faster training than the original one. In comparison with the multilayer perceptron, the dynamic programming technique and hidden Markov models, we empirically demonstrate that the IOHMM architecture is a promising way to text-dependent speaker identification.