A Novel English Speech Recognition Approach Based on Hidden Markov Model

In order to promote English teaching quality, English speech recognition has been attracted more and more attention. In this paper, we aim to propose a novel English speech recognition approach based on Hidden Markov model. Particularly, the English speech recognition problem can be described as seeking the most suitable word sequence based on a segment of English voice. Furthermore, using the Hidden Markov model, English speech recognition problem can be converted to find a word sequence, which is translated to a sequence of Hidden Markov model. To promote the performance of the standard Hidden Markov model, we propose a HMM-based semi-nonparametric method to enhance the performance of the accuracy of English speech recognition. Probabilistic transition frequency profile matrix and average probabilistic emission matrix are calculated for all training word sequences. Experimental results demonstrate that the proposed English speech recognition approach can achieve higher recognition rate than the related works.