Vision-Based Driver Drowsiness Detection: From Deep Learning Models to Real-Time Mobile Deployment

A significant percentage of traffic accidents worldwide result from sleepy drivers. Although detection methods have been established, their utility is often problematic. Physiological signals (EEG, ECG) and vision-based behavioral cues (eye closure, yawning) have been studied extensively, and deep learning models such as CNNs have shown excellent accuracy in controlled settings. Significant gaps still exist, however, especially regarding robustness against varying lighting and occlusions, on-road validation, and computationally efficient non-intrusive systems for real-time mobile deployment. This paper synthesizes and critiques current vision-based approaches and proposes a lightweight CNN architecture (MobileNetV2) optimized for ondevice inference with TensorFlow Lite, offering a scalable solution for road safety on common Android devices.