This research study deals with the crucial problem of driver drowsiness, which is the top reason for road accidents in the world. The system comprises a deep learning mechanism that tracks the face and eyes of the driver in real-time using a webcam. Employing MobileNetV2 architecture in combination with MRL dataset, this system can predict whether the eyes of a driver have been closed for some considerable amount of time so that an alarm can be generated through a buzzer. If the driver rejects many warnings, the system combined with a car simulation that obviously slows down the car and moves to another lane provides for maximum safety. From the initial observations, the model has shown very good accuracy in telling open from closed eyes, so creating an actual real-time drowsiness detection method is effective. This work demonstrates where such technology could be installed in vehicles to ensure better safety on roads and reduce the risk posed by driver drowsiness.
Driver Drowsiness Detection Using MobileNetV2 and Deep Learning
2024-12-11
1010779 byte
Conference paper
Electronic Resource
English
Driver Drowsiness Detection Using Deep Learning
IEEE | 2023
|Driver Drowsiness Detection Using Deep Learning
IEEE | 2021
|Driver Drowsiness Detection using Deep Learning
IEEE | 2022
|