Smart Facial Features for Real-Time Operator Sleepiness Prediction Using Hybrid Convolutional Neural Network in Computer Vision

Pooja S , Subash V , Mahesh G

Department of Information Technology, Agni College of Technology, India

IJTCSE-ISSN 2349–1582

Volume No :10, Issue:02

Accepted for June 2023 Issue

Abstract

Operator Sleepiness prediction is a process of detecting when an Operator is experiencing Sleepiness or fatigue while driving a vehicle. This is an important safety feature, as Sleepy driving can lead to accidents and injuries. There are several methods used to predict Operator Sleepiness, including physiological monitoring, behavioural monitoring, and hybrid methods. Physiological monitoring methods like CNN involve measuring the Operator’s physiological signals, such as image or video frame processing from a camera. These frames can provide information about the Operator’s level of alertness and can be used to detect Sleepiness. Behavioural monitoring methods that are DCNN on the other hand, involve observing the Operator’s behaviour, such as comparing the frames with the processed dataset. This information is mainly used to detect Sleepiness. Hybrid methods combine physiological and behavioural monitoring methods of CNN and DCNN and added to the fuzzy logic algorithm makes an HDCNN (Hybrid Deep Convolutional Neural Network) to provide a more comprehensive assessment of the Operator’s level of Sleepiness and improves the accuracy.