<![CDATA[Driving while drowsy is identified as a significant risk factor in traffic accidents, yet awareness of this risk is often lower compared to other hazards. Phenomena such as microsleep have been shown to increase the risk of inattention and accidents on the road. This study proposes a novel approach utilizing Deep Learning, specifically YOLOv8, to detect and address the risk of driver drowsiness. To train the model, the researchers employed a secondary dataset consisting of 3708 images, partitioned into 80% for model training and 20% for validation. Multiple models were compared during the training process, and the results indicated that the YOLOv8 model outperformed previous models, achieving a recall value of 0.95261, precision of 0.94655, F1-SCORE of 0.9496, and mAP of 0.98055. This research contributes to the development of more effective drowsiness detection systems using Deep Learning approaches, with promising evaluation results.]]>
Copyrights © 2023
