I would like to know more about the AI CoPilot sample you provided if possible.
1- Do you use facial landmark detection for gaze tracking and drowsiness detection, or another approach?
2- Do you create a profile information for each driver to track their behavior according to their average features? If so, which information will be stored for each driver?
3- Do you perform yawning detection to help detect drowsiness or do you use only eye/head positions?
Maybe some of the questions are too detailed for you to answer, but I would really appreciate it if you could give me some insight.