Generation and perception of F₀ markedness for communicative speech synthesis

Aiming at natural F₀ control for conversational speech synthesis using attributes of constituent output words, F₀ characteristics are analyzed from both generation and perception viewpoints. We recorded commonly used two-phrase utterances consisting of Japanese adjective and adverb phrases expressing different degree of markedness under designed conversational situations, and compared their F₀ characteristics. The comparison showed the consistent F₀ control dependencies not only on adverbs themselves but also on the attribute of following adjective phrases. Strong positive or negative correlation is observed between the markedness of adverbs and F₀ height when an adjective phrase showing positiveness or negativeness is followed to the current adverb phrase. These consistencies have been perceptually confirmed by naturalness evaluation tests using the same two-phrase samples with different F₀ heights. Finally, a computational model of conversational F₀ control is proposed using lexical information of adjectives showing positiveness or negativeness and adverbs expressing markedness. F₀ estimation experiments quantitatively showed the possibility of F₀ control for natural conversational speech synthesis using the attribute of constituent output words.