So I started a little Hack-A-Thon project over Thanksgiving. I initially was just connecting a single ReSpeaker Mic Array V.2 to a RaspPi running Ubuntu Mate and went through the tutorials. I also looked at a project called ODAS which had some pretty cool looking visualizations of Direction of Arrival and Angle of Arrival of detected sound. Unfortunately their visualization does not really run well on the PI. But... I am able to run part of the project on the PI and push the sound localization data to port on another computer. Also, I really don't need the visualizations. They are nice to know it is working but Junior is not going to use the visualization just the backing data. 

It was a bit of a challenge to make sure that when connecting two identical devices to the PI that I could keep the two straight. I ended up adding a little Python code to determine which physical USB port on the PI was being used to designate the connection left and right. I also had the same issue with running two instances of the ODAS code for localization. I am still working on expanding my python script to modify the config files that are used when launching the two instances of odaslive so that I can determine the correct configuration file for left and right microphone arrays.

The next piece I need to work out is passing the audio stream from the PI to another computer to do the Speech to Text conversion. I am thinking I SHOULD be able to do this with running MRL on both the PI and the other computer but we will see if I can remember how all that works. 

For now here is a beautiful picture of my desk with scrolling data in terminal windows...