Real-Time Drowsiness Identification Based on Eye State Analysis using Efficientnetb7 Deep Learning Technique

Abstract

The amount of road accidents caused by driver drowsiness is one of the world’s major challenges. It is critical to prevent catastrophic accidents and reduce the financial burden on society caused by driver drowsiness. In these occurrences, this research endeavors to optimize efforts towards the real-time identification of drowsiness in drivers under authentic driving conditions, with the overarching objective of mitigating the incidence of traffic accidents.

We propose a novel system aimed at alerting drivers to instances of distraction or drowsiness. Image processing techniques and Convolutional neural networks (CNNs) based on EfficientNetB7 version is employed in real-time applications. the camera continuously monitors the driver's facial features and eye movements, with a particular emphasis on eye tracking as a key indicator of drowsiness. We used Dlib toolkit to provide a new, more stable parameter for evaluating the status of the driver’s eyes.

The database used is Driver Drowsiness Dataset (DDD), which contains image of drivers with different levels of drowsiness. The proposed model was evaluated in terms of accuracy, precision and Loss in detecting drowsiness in the eye region. The results of the study show that the EfficientNetB7 model have high accuracy in detecting drowsiness in the eye region.