Edge-AI Driver Drowsiness & Distraction Telemetry on Jetson Nano / ESP32-CAM
Real-time edge vision system calculating Eye Aspect Ratio (EAR) and head pose angles at 30 FPS to trigger instant cabin buzzer alarms.
Project Overview
Deploys an optimized MobileNetV3 and facial landmark detector onto edge hardware (NVIDIA Jetson / Raspberry Pi). Computes real-time PERCLOS (percentage of eyelid closure), yawning frequency, and phone distraction, sounding alerts before micro-sleep accidents occur.
Deploys an optimized MobileNetV3 and facial landmark detector onto edge hardware (NVIDIA Jetson / Raspberry Pi). Computes real-time PERCLOS (percentage of eyelid closure), yawning frequency, and phone distraction, sounding alerts before micro-sleep accidents occur.
Real-time edge vision system calculating Eye Aspect Ratio (EAR) and head pose angles at 30 FPS to trigger instant cabin buzzer alarms.
Core Project Objectives
Capture continuous analog/digital sensor readings with robust noise filtering and hardware calibration.