From last week, I got the “base” rotation working. I implemented it by having 2 servos moving in the same rotation axis. I also added in Anna’s solution for the microphones. From there we continued to fine tune integration.

This week we were finalizing out integration. We ran into some bugs with latency, audio detection, and motor movement. Edward managed to fix the latency by checking for facial detection only if there is audio difference. Eddy and I also paused video and audio whenever the motors were moving to remove any motor movement from audio calculations and remove the shaky video clips. I also worked with Eddy to debug the implementation to crop frames if the user is out of range for the servos to center. With audio detection, Anna and I brainstormed optimizations because the motor noises were previously included in calculations, and sometimes the microphones picked up noises way out of range. Anna implemented a rolling median solution and I tested. Lastly, we’re having issues with the stepper motor moving really slowly. I reasoned this is because the stepper movement happens 1 command at a time, and now that all the code is in, the other threads are cutting into this. I made changes such that all other threads pause execution when the stepper motors are moving, and this works now.