Quantitative analysis of F0 contours of emotional speech of Mandarin

For emotional speech synthesis, a quantitative model giving a parametric representation of F0 contours is needed. Purpose: investigate quantitatively F0 characteristics of Mandarin speech in four basic emotions (anger, fear, joy, and sadness) and in neutral reading. Two approaches are compared: surface features analysis from time-normalized F0 contours analysis-by-synthesis of time-intact F0 contours based on the command-response model Model-based analysis-by-synthesis of F0 contours Tone commands in neutral read speech: each syllable should basically comply with the inherent command pattern for the particular tone type. in emotional speech: much more variation, but generally, the polarity of a tone command is rarely reversed. stressed syllables preserve tone command patterns better. Phrase commands largely aligned with major syntactic boundaries can be determined by comparison of syllabic F0 patterns and the canonical form of tones, by comparison of F0 patterns in adjacent tones, and also with the aid of prosodic perception. only assigned when necessary and linguistically meaningful Sentential F0 declination: the magnitude ranks as follows: anger > joy > neutral > fear > sadness, which also indicates active emotions > neutral > passive emotions.