Fourth ISCA ITRW on Speech Synthesis
August 29 - September 1, 2001
The ACTORŽ Text-To-Speech (TTS) synthesis system, developed at Loquendo S.p.A., is here described. The system employs a unit-selection concatenative synthesis technique, relying on labeled acoustic databases providing phonetic and prosodic coverage of the intended language/domain and on an original algorithm for run-time selection of the acoustic units to be concatenated. This technique yields high-naturalness and human sounding voices. ACTORŽ is a multi-voice and multi-language system, exploiting different kinds of language dependent knowledge (grammatical, phonetic and prosodic, as well as acoustic) with the support of several development tools (statistical tools for database design, machine learning algorithms, tools for speech signal analysis and phonetic alignment, etc.).
Bibliographic reference. Quazza, Silvia / Donetti, Laura / Moisa, Loreta / Salza, Pier Luigi (2001): "ACTOR: A multilingual unit-selection speech synthesis system", In SSW4-2001, paper 209.