Method for analyzing speech involving detecting the formants by division into time frames using linear prediction

Abstract

A process for speech analysis and more specifically an automatic process for the analysis of continuous speech. The waveshape of the speech is described with the aid of the resonant frequencies, formants, which arise in the speech organ. The process determines suitable frequencies for the formants from an utterance by dividing the utterance into time frames and analyzing the utterance by linear prediction in order to determine roots of the denominator polynomial and thereby frequency values for each frame. The utterance is divided into voiced regions and in each voiced region the centers of vowel sounds are established in order to obtain a number of starting points. Tracks are formed from the starting points by sorting the roots from frame to frame so that old and new roots are linked together. Factors of merit are calculated for the tracks relative to formants and the tracks are distributed to formants in accordance with the factors of merit. The factors of merit take into consideration the bandwidth, continuity and relation to the formants of the tracks. The process gives a global optimisation by delaying the formant allocation until a complete voiced region has been analyzed. By linking the tracks together in this way, additional/false resonances can be controlled, which resonances arise in association with linear prediction.