EULER Language Provider's Guide

This section assumes you are somehow familiar with the general organization of a TTS system and and with that of its components. If not, you might want to read this introductory paper on TTS synthesis, and maybe buy a good book on TTS synthesis (why not this one?).

Incorporating a new language into EULER requires to either build new modules for its preprocesing, morpho-syntactic analyzer, phonetizer, prosody generator, and speech synthesizer, or to customize existing EULER modules (assuming they are somehow language-independent, which is the case for most existing EULER modules). Customization itself can be achieved by either using the source code of existing modules and adapting it (thereby creating a new module), or by creating new databases for an existing module. This last solution is by far the simplest; we recommend it whenever possible.

The first and most obvious thing you can do is to take a look at the documentation of the EULER modules for French. You will see the features of these modules and how they share data through the MLC.

Then try to figure out whether availble modules can be directly reused (you should be able to do this without having to examine their source code). To do this, you will have to clearly understand what modules do and how. Pay special attention to the database they use, and try to figure out what would have to be adapted for your language.

We are currently working on a simple example for building an American English and an Arabic TTS for EULER, and will provide all sources for it as soon as it is done. These examples will show how to quickly incorporate a new language into EULER.


Copyright © 1999 TCTS LAB, Faculté Polytechnique de Mons, Belgium