Hélène Valbret
Modelling speaker-specific intonation
characteristics
Abstract:This report describes work carried out at ATR during my post-doctoral year, to model the speaker-specific characteristics of intonation using information extracted from the fundamental frequency
of the speech waveform. It also describes an implementation of pitch-synchronous signal processing techniques for realisation of intonation
in the CHATR speech synthesis system, and the tools developed for
the extraction, analysis, and resynthesis of fundamental frequency information.