I'm interested in implementing the MARY TTS engine on Android, specifically to use some of the advanced features available to the voices (such as inflection, duration, pitch, etc). Ideally I would create a self-contained Android application with a front-end (not MARY) to handle the text input and generation, and a back-end (is MARY) to handle the Text-to-Speech task and pass the audio files back to the front-end.
Has anyone attempted this? Where would be a good place to start?
We have already start to solve this task with integrating MaryTTS java code to Android. Can tested from here