Skip to content
02/21
- We checked that we are now able to ssh into Sandy as well. So now it is possible to do:
- ssh pi@comovosam.wifi.local.cmu.edu
- ssh pi@comovosandy.wifi.local.cmu.edu
- We began discussing and researching to find answers to the design questions that we came up with and noted down earlier this week.
- We looked into PyAudio for real-time microphone signal processing and OpenCV for gesture detection (for manual mode) from a camera video feed. Any microphone (connected via USB) should work and any camera (connected to the camera jack) should work.
- We are going to do sound triangulation (most likely without an open-source library) using 3 microphones and also use the OpenCV video feed method to detect faces/heads and aid us in this process. Microphones probably shouldn’t rotate with device, but camera should.
- Need to build ML models to classify “hand/not hand” for gesture detection and “face/not face” for use in sound localization part.
- We explored using a motor HAT to make interfacing with the motor easier. Stepper motor seems to be a good motor choice. Looking into adafruit_motorHAT library (very user friendly functions and no need to directly communicate with GPIO pins).
- Need to finalize exactly which models of motor, motor HAT (?), camera and microphone we want to use.
- Started drawing block diagram and making slides for design review presentation next week.