Advancements in Driver Safety using RESNET101 for Real-Time Drowsiness Detection

In recent years, many professions require longer focus time. Any drivers must constantly be aware of their surroundings on road to react appropriately to any unexpected occurrences. Sleepy drivers cause many road accidents and traffic mishaps. So, we need a strong system that can identify the unwellness of a driver and inform them about their physical capability and mental capability. This might really reduce the number of sleepy driving crashes. Even so, there are numerous challenges that stand in the way of developing such systems when it comes to effectively and efficiently detecting driver fatigue symptoms. A vision-based technology is one way to operate drowsiness detection systems. This website describes the technologies used in the detection of driver drowsiness. The specifics of utilizing the vision framework to ascertain whether a driver is drowsy are also investigated. To detect drowsiness, the proposed approach monitors the situation of eyes of the driver, extract the eye aspect ratio of both the eyes. If there is any drowsiness in drivers' eyes then, an alarm gets sounded. Analysis of the dataset suggests work has a 96% accuracy using ResNet-101 model on a substantial subset of the MRL eye dataset consisting of pictures of eyes.