Speech Prosody 2006

Dresden, Germany
May 2-5, 2006

Language- and Speaker Specific Implementation of Intonation Contours in Multilingual TTS Synthesis

Boris Lobanov (1), Liliya Tsirulnik (1), Dmitry Zhadinets (1), Helena Karnevskaya (2)

(1) United Institute of Informatics Problems N.A.S., Belarus
(2) Minsk Linguistic State University, Belarus

The paper is concerned with the study of final/non-final phrase intonation and its language- and speaker-specific peculiarities. A phrase, according to the model used, is represented by a sequence of accentual units consisting of pre-nucleus, nucleus and post-nucleus. Experimental data obtained from the prosodic analysis of a text, recorded by Russian and Polish native speakers, has made it possible to create accentual unitsí "portraits" for different types of final/non-final phrase intonation. The implementation of these "portraits" in the unified text-to-speech synthesis system for Slavonic languages with the ability of personal speaking manner cloning is discussed.

Full Paper

Bibliographic reference.  Lobanov, Boris / Tsirulnik, Liliya / Zhadinets, Dmitry / Karnevskaya, Helena (2006): "Language- and speaker specific implementation of intonation contours in multilingual TTS synthesis", In SP-2006, paper 159.