I see that OpenAI has open-sourced (MIT license) their Whisper voice recognition technology. I wonder if this would make it possible to create a voice-activated local controller for Roon?
Anyone tried it out yet?
I see that OpenAI has open-sourced (MIT license) their Whisper voice recognition technology. I wonder if this would make it possible to create a voice-activated local controller for Roon?
Anyone tried it out yet?
This might be an interesting starting point. The developer takes the audio, runs it through Whisper, then fuzzy-matches it to a command from a specified set of commands.
Sounds scary I’m waiting for the moment I’ll record Whisper so to mach that to a set of my movements
Mostly in Python as well . Thanks for guiding me to a warren full of potential rabbit holes.
It will play Galactic if you Boogaloo, it will play B52’s if you do the Swim.
Actually, that sounds kinda cool.
In this case I wander what Whisper thinks about Snake Oil
I realize it may be hard to tell with me, but this is not a whimsical post. Hard-edged technical. I’m feeling the need to have voice control of my Roon setup, and given some good starting points, like the Roon Command Line interface, and good speech recognition technology, like Whisper, I may have to build my own.
100% agreed. As a not native English speaker none of the speech recognition the I played with did worked. The fun part was to the general AI “madness”, I do digital drawing/illustration (my profile pic is a digital painting) and in the last year the AI topic is a hot one.
I did not mean to be impolite.