An IOT and Machine Learning-Based System for Detecting Driver Drowsiness

Improving road safety is still a top priority, since sleepy driving has been found to be a major risk factor for collisions. This research offers a novel real-time visual-based sleepiness detection method designed specifically for drivers. Using a drowsiness-focused dataset, our approach finds and separates the face region in every frame, then uses a strong facial landmarks detector to extract the eye region accurately for analysis. Then, as a critical measure of drowsiness, the eye aspect ratio is calculated and tracked over frames. Two different models—Inceptionv3, and YOLO—are combined with transfer learning approaches to increase detection accuracy. By classifying eye condition and distinguishing between closed and open eyes, our system can detect extended eye closure and sound an alarm to warn the driver of possible drowsiness. This research offers a proactive way to reduce sleepiness behind the wheel and emphasize road safety by providing a dependable and timely strategy to mitigate it.