Real Time Driver Drowsiness Detection and Alert System

Instant Driver Drowsiness Detection and Warning System helps detect whether a person is drowsy or not. We trained VGG16 and a simple CNN model to detect fatigue and lack of sleep. The system uses the VGG16 architecture for feature extraction and the CNN model for classification, aiming to accurately identify the symptoms of a drowsy driver. We also used the Haar Cascade Classifier to train the model and check sleep and insomnia rates. When detected, the system activates an alarm mechanism, i.e. an audible signal, to immediately notify the driver and reduce the risk of accidents due to drowsy driving. When detected, the system activates an alarm mechanism, i.e. an audible signal, to immediately notify the driver and reduce the risk of accidents due to drowsy driving. The incorporation of eye tracking technology augments the system's capability to precisely evaluate the driver's mental condition in real-time, offering a pre-emptive strategy to alleviate the hazards linked to impaired driving. The results show that the VGG 16 model outperforms the convolutional model with an accuracy of 96.3%.