Real-Time System for Driver Fatigue Detection Based on a Recurrent Neuronal Network.

Younes Ed-Doughmi,Youssef Hbali,Najlae Idrissi

doi:10.3390/jimaging6030008

Younes Ed-Doughmi, Youssef Hbali + Show 1 more

Open Access

https://doi.org/10.3390/jimaging6030008

Copy DOI

Abstract

In recent years, the rise of car accident fatalities has grown significantly around the world. Hence, road security has become a global concern and a challenging problem that needs to be solved. The deaths caused by road accidents are still increasing and currently viewed as a significant general medical issue. The most recent developments have made in advancing knowledge and scientific capacities of vehicles, enabling them to see and examine street situations to counteract mishaps and secure travelers. Therefore, the analysis of driver’s behaviors on the road has become one of the leading research subjects in recent years, particularly drowsiness, as it grants the most elevated factor of mishaps and is the primary source of death on roads. This paper presents a way to analyze and anticipate driver drowsiness by applying a Recurrent Neural Network over a sequence frame driver’s face. We used a dataset to shape and approve our model and implemented repetitive neural network architecture multi-layer model-based 3D Convolutional Networks to detect driver drowsiness. After a training session, we obtained a promising accuracy that approaches a 92% acceptance rate, which made it possible to develop a real-time driver monitoring system to reduce road accidents.

Highlights

The World Health Organization (WHO) has identified road traffic injuries as a major global public health problem
We propose a variation of the Recurrent Neural Networks (RNN) calculation that has just demonstrated its effectiveness on drowsiness detection, which is distinguishing an activity-dependence on a sequence
Models consist of logistic regression [36], artificial neural network (ANN) [35,38], support vector machine (SVM) [39], hidden Markov model (HMM) [8], Multi-Timescale by CNN [15], long-short term memory (LSTM) network smoothed by a temporal CNN [4], or end-to-end 3D-CNN [6] are conducted

Summary

Introduction

The World Health Organization (WHO) has identified road traffic injuries as a major global public health problem. A study conducted in Morocco [2] in 2013 demonstrated a worrying number of accidents identified with drowsiness, which is responsible every year for more than 4000 deaths and 1.4 billion dollars of material harm. Deep learning is an upgrade of artificial neural networks, made up of a more layered structure allowing higher levels of abstraction and better predictions from the data [33,34]. Models consist of logistic regression [36], artificial neural network (ANN) [35,38], support vector machine (SVM) [39], hidden Markov model (HMM) [8], Multi-Timescale by CNN [15], long-short term memory (LSTM) network smoothed by a temporal CNN [4], or end-to-end 3D-CNN [6] are conducted.

Objectives

Methods

Results

Conclusion