Tutorial at INTERSPEECH 2007, Antwerp, Belgium
Voice Transformation refers to the various modifications one may apply to the sound produced by a person, speaking or singing. A major application area of Voice Transformation is that of concatenating speech synthesis. Voice Transformation is a flexible, could be a simple, and efficient way to bring the variety needed in the current Text-To-Speech synthesis systems based on concatenation of units (large or small). Recently, because the speech quality produced by concatenating speech synthesis systems has considerably been improved, the demand for high quality Voice Transformation algorithms grew a lot.
In this tutorial we will provide a description of various ways in which someone can modify voice and provide details on how to implement these modifications using a simple, although quite efficient, parametric model based on a harmonic representation of speech. Discussing quality issues of current voice transformation algorithms in conjunction with properties of the speech production and perception systems we will try to pave the way for more natural Voice Transformation algorithms in the future.
Presenter
Yannis Stylianou
University of Crete, Computer Science Dept.
Short Bio
Yannis Stylianou is Associate Professor at University of Crete, Department of Computer Science since 2002. He received the Diploma of Electrical Engineering from NTUA, Athens, Greece in 1991 and the M.Sc. and Ph.D. degrees in Signal Processing from ENST, Paris, France in 1992 and 1996, respectively. From 1996 until 2001 he was with AT&T Labs Research (Murray Hill and Florham Park, NJ, USA) as a Senior Technical Staff Member. In 2001 he joined Bell-Labs Lucent Technologies, in Murray Hill, NJ, USA. He was Associate Editor for the IEEE Signal Processing Letters. Currently is Associate Editor for the EURASIP Journal on Speech and Audio Processing and Vice-Chairman of the COST Action 2103: “Advanced Voice Function Assessment” developing algorithms for voice quality control. He holds 8 patents.