4D: A Real-Time Driver Drowsiness Detector Using Deep Learning

There are a variety of potential uses for the classification eye conditions, including tiredness detection, psychological condition evaluation, etc. Because its significance, many studies utilizing typical neural network algorithms have already been published in literature, with good results. Convolutional networks (CNNs) employed real-time applications to achieve two goals: high accuracy and speed. However, identifying drowsiness at an early stage significantly improves chances being saved from accidents. Drowsiness detection can be automated by using artificial intelligence (AI), which allows us assess more cases less time lower cost. With help modern deep learning (DL) digital image processing (DIP) techniques, this paper, we suggest CNN model state categorization, tested it on three models (VGG16, VGG19, 4D). A novel named 4D was designed detect based state. The MRL Eye dataset used train model. When trained training samples same dataset, performed very well (around 97.53% predicting test dataset). outperformed performance other pretrained VGG19). This paper explains how create complete system that predicts driver’s eyes further determine drowsy alerts driver before any severe threats road safety.