A Visual Fusion Based Deep Learning approach for Real Time Driver Drowsiness Detection
Main Article Content
Abstract
Growing concerns due to increased road accidents due to driver drowsiness have paved the way for the development of real time driver sleepiness detection systems.The real world implementa -tion of these systems can play a major role in reducing the accidental rate due to driver sleepin -ess. This paper suggests a fusion based approach for identifying the driver’s sleepiness in real-time.The model uses driver’s behavioral features like eye state and yawning as a metric for sleepiness identification. Mouth Aspect ratio is used for yawn detection and Convolution Neural Network for eye state classification. Dlib’s HOG based detection is used for facial feature detection in this paper. Our proposed model shows a high accuracy rate and can detect if the driver is drowsy or not. It performs well in various lighting conditions and also if the driver wears spectacles. Training and testing accuracy of the proposed CNN model is 98.75% and 97.65% respectively. Moreover, it is found that our proposed methodology performs well in real time.