On fundamental frequency contour synthesis and control method for Chinese Speech Synthesis

Based on Chinese tone characteristics and knowledge of phonetics, improving on Fujisakipsilas command-response model, generation model of fundamental frequency contour for Chinese Speech Synthesis is presented to simulate intonation, phrase tone trend and prosodic change of syllable for Chinese utterances. Under controlling of phrase command and tone command, which reflects all the relationships of prosodic hierarchy and prosodic parameters, simple methods have been proposed for model parameters setting and F₀ synthesis based rules. The experimental results show the proposed methods to be available in this paper.