PHONEME RECOGNITION IN CONNECTED SPEECH USING BOTH STATIC AND DYNAMIC PROPERTIES OF SPECTRUM DESCRIBED BY VECTOR QUANTIZATION.

The authors describe an approach to phoneme recognition based on a clustering method which considers phonemic featuers in each frame. In the clustering, both acoustic and phonemic features of speech are used. The acoustic features are linear predictive coding (LPC) coefficients, the cepstral changes between adjacent frames, and the power changes. The combination of these features shows both the static and dynamic properties of the spectrum. The phonemic feature in a frame is composed of a triplet of phonemic symbols. A vector quantization method is used for the clustering. An experimental extraction of phonemic label sequences is performed, considering a contiguity of code sequences between input and the reference phonemic patterns. 8 refs.