I’m trying to make a small voice controlled device with a raspberry (kind of homemade alexa / google home) and i was thinking on recording the voice with your library.
I’d like to have a fixed size buffer (for example 3 seconds of sound) where i can analyze and detect that somebody has said something before sending it to the corresponding service.
Is it posible to store the audio recorded on a kind of circular buffer so i can do something similar to my idea or does it sound like something a bit awful?
I’m open to all kind of ideas.
Thanks a lot,