I am seeking advice for launching a vocal assistant targeting african dialects. The purpose is for educational services.
What sould be the starting point of such project? The speech recognition engine? Voice dataset?
The project is at very early stage. Developers expert in the domain are more than welcome
Hi @maseyan, I think the voice data set is a good place to start, as you will need some data to make a speech recognition engine.
Have you seen this project: http://alffa.imag.fr/
^^ Working with them might be a way to have a high impact and not start from scratch.
Hello Michael,
Thanks so much for your response. What kind of deal could I strike with them?
Not sure 
Very soon we are going to put Common Voice on Pontoon, and then you can translate into any language you want.
Hello @maseyan,
I am also interested in ASR for African (West African, in particular) languages.
If you go to https://voice.mozilla.org/en, there is “Request a Language” at the bottom, where potential contributors can express their interest to contribute in their language (e.g. Wolof, Fula). So if you know people that know those languages, please tell them they can express their interest there…