AUTOMATIC PHONE SEGMENTATION AND LABELING OF CONTINUOUS SPEECH

To obtain an accurate phone sequence from a continuous speech signal, we suggest a novel approach consisting of tightly coupled bottom-up an d top-down processing. The bottom-up path consists of segmentation, re cognition and labeling. Also the top-down path consists of labeling, s peech generation and segmentation. In this manner, the four processes form a closed feedback loop achieving an optimal interpretation effici ently for a given noisy observation of speech signal and a priori know ledge. The major goal of this paper is to identify the system model us ing both the stochastic estimation theory and the mean field theory. E xperimental results are obtained in terms of the TIMIT database. It is shown that introducing the top-down path to the traditional bottom-up path can improve the recognition rate by 19.7%, and reduce the error (substitution, deletion and insertion) rate by 16.1%. As a result, the overall system can transform the incoming continuous signal into one of the 61 phone classes at the rate of 73.7%.