Open Martin2kid opened 4 years ago
Hey, just wanted to give you an update on this issue.
I'm experimenting with a 5/8 radial mic array now. Will keep you posted if I publish my directional models. I think the scope of those models is a bit beyond this project, so I'd release in a separate codebase.
This is just future suggestion to already impressive work,
What if rpi-deep-pantilt can also triangulate audio source, point to that direction.
Perhaps 2 microphone practically spaced far enough apart & measure 2 audio channel input speed difference, triangulate & apply toward servo angle processing and point to that direction? (I know it'll perhaps require 360 degree servo)
https://www.tensorflow.org/api_docs/python/tf/audio/decode_wav