Title :
Isolated word recognition using interframe dependent hidden Markov models
Author :
Ming, J. ; Smith, F.J.
Author_Institution :
Sch. of Electr. & Comput. Eng., Queen´´s Univ., Belfast, UK
Abstract :
A new hidden Markov model (HMM) with first-order dependent observation densities is presented to account for the statistical dependence between successive frames. In this model, the dependence relation among the frames is optimized to maximize the likelihood for both the training and testing data. Experimental comparisons with the standard continuous density HMM as well as the conditional-observation HMM for an isolated word recognition task show that a significant performance improvement is achieved for the new model.<>
Keywords :
hidden Markov models; speech recognition; statistical analysis; conditional-observation HMM; continuous density HMM; first-order dependent observation densities; interframe dependent hidden Markov models; isolated word recognition; performance; speech frames; statistical dependence; testing data; training data; Binary search trees; Computer science; Density functional theory; Hidden Markov models; Industrial economics; Integrated circuit modeling; Parameter estimation; Speech; State estimation; Testing;
Journal_Title :
Signal Processing Letters, IEEE
