Efficient Viterbi scoring architecture for HMM-based speech recognition systems

Abstract

A new dedicated architecture for Viterbi scoring in hidden Markov model (HMM)-based real-time isolated word recognition systems is proposed. Because, in HMMs, most states are connected to only three or fewer preceding states, the state transition matrix is a banded upper triangular matrix. Using this property of the HMM, we design an efficient Viterbi scoring architecture.