Kalliope - modular voice controlled
I have tried several voice control modules and none completely convinces me.
I am also testing in Linux with a project that could be very interesting if it could be adapted to Magic Mirror. It is fast and efficient, and with Wit.ai (STTS) and pico (TTS) there are no time limits, with an efficiency of 100% in accuracy.
It’s called Kalliope: [https://github.com/kalliope-project/kalliope](link url)
Try it independently, it will surprise you