Would real-time speech recognition be feasible on a raspberry pi 4 (4GB) overclocked to 2.0Ghz?
No need to overclock, switching to TFLite engine will do the trick. We’re working on that right now.
1 Like
That’s awesome! Any ETA on when it will be stable enough to use?
Define what you mean by “stable enough to use” ?
Is there a tutorial on the github wiki which explains how to set it up on a pi?
There’s nothing specific to do on RPi, there are already plenty of docs on how to use.