Create Speech recognition plugin

[lawrab]

Feature request.

Add a speech recognition plugin that can do keyword matching and raise events when keywords are recognized in an audio stream.

There are cloud services available for this, however, this would use bandwidth which could impact on streamers network usage and needs a bit of testing,

AWS, Azure and Google offer speech recognition services.

It might be worth looking at tensorflow so users can train their own keywords, not sure, not done this before.

https://www.tensorflow.org/tutorials/audio/simple_audio

[lawrab]

Can look at https://github.com/SciSharp