Driver Monitoring based on Drowsiness and Yawning using YOLOv8

The Samruddhi Mahamarg, a vital expressway, witnesses numerous accidents each year due to driving hypnosis and drowsiness, especially on its long stretches. This thesis proposes a novel approach to enhance safety by addressing these issues through the identification of drowsy eyes and yawning using state-of-the-art deep learning models. The system employs two YOLOv8 models, one dedicated to detecting drowsy eyes and the other for recognizing yawning behavior. Remarkably high accuracy rates of 99.92% for drowsy eyes and 99.95% for yawning demonstrate the robustness and reliability of the proposed models. By accurately identifying these critical signs of driver fatigue, the system provides timely alerts, mitigating the risks associated with driving hypnosis. This research emphasizes the importance of mitigating driving hypnosis to reduce the frequency of accidents on highways like Samruddhi Mahamarg, emphasizing the potential to save lives and prevent injuries. The incorporation of cutting-edge technology in real-time monitoring showcases the commitment to advancing road safety in high-risk environments.