This package holds multiple ways to generate speech by text:
- Mary TTS http://mary.dfki.de/
- flite: A small fast runtime synthesis engine
- pico2wave: This uses an android speech engine
- More to come
We have tried multiple approaches that have different favors and bottlenecks.
First Mary TTS generates good sounding speech and has an easy interface for installing new voices but consumes a lot of memory. Therefore it is not suitable for the use on single board computers as of now.
Head over to http://mary.dfki.de/download/index.html and grab an installer. The mary tts node assumes that you have started the mary tts server beforehand on the standard port.
This is a small and fast version of festival. However, with the voices and params we tried speech generation is still to slow for single board computers.
sudo apt-get install flite
This tool speech synthesis engine is really fast for usage on single board computers. There are not many parameters existing but some languages as english and german.
sudo apt-get install libttspico-utils
Mary TTS bibtex:
@article{schroder2003german,
title={The German text-to-speech synthesis system MARY: A tool for research, development and teaching},
author={Schr{\"o}der, Marc and Trouvain, J{\"u}rgen},
journal={International Journal of Speech Technology},
volume={6},
number={4},
pages={365--377},
year={2003},
publisher={Springer}
}
Flite bibtex:
@inproceedings{black2001flite,
title={Flite: a small fast run-time synthesis engine},
author={Black, Alan W and Lenzo, Kevin A},
booktitle={4th ISCA Tutorial and Research Workshop (ITRW) on Speech Synthesis},
year={2001}
}