next up previous
Next: RULAMO: Rule-Based Language Model Up: Some projects of the Previous: VoiceTools: Methods and Tools


PROSO: Prosody Control for Polyglot Speech Synthesis

The long-term objective of this project is a TTS synthesis system that will be able to process any mixed-lingual German, French, Italian, and English texts. The specific topic of this project is prosody. We first have to do research into fundamentals of prosody of mixed-lingual speech, in particular into the prosody of foreign inclusions uttered by professional speakers. With the results of this work we will build an improved prosody control for our PolySVOX TTS synthesis system (see project PolyTTS).

Publications related to this project are [Rom09a], [Rom09c], [HP10], [HP12] and [HP13]. Additionally, the following reports belong to this project: [Hof09] and [Hof10].

Supported by: This project is mainly supported by ETH Zurich.

Last updated: Mon Nov 20 15:00:45 CET 2017 by: Beat Pfister