System for Alertness and Fatigue Estimation Using Deep Recurrent Intelligence for Vehicle Engagement

View at publisher Summarise with VIRAT
Authors: Panigrahi Jyothsna, Sarita Panda, Lanka Sri Vidhya Lavanya, Karri Harshitha, Sanapala Pavan Kumar, Byri Leela Suryakala
Cited by: 0
References: 12

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Driver drowsiness remains a critical factor in road traffic accidents, accounting for substantial casualties worldwide. This research presents an automated drowsiness detection system utilizing convolutional neural networks and real-time video processing. The system monitors eye states through webcam input, classifying them as open or closed using a trained deep learning model. A Haar Cascade classifier detects facial regions, followed by eye extraction and classification. The implemented VGG16 model achieves 97.62% testing accuracy with a compact architecture suitable for real-time applications. Alert mechanisms activate upon detecting sustained eye closure for 15 consecutive frames, providing both visual and auditory warnings. The system demonstrates robust performance under controlled environments, offering a practical solution for preventing fatigue-related incidents in transportation and occupational settings.

Fields of Study & Subjects

Sleep and Work-Related Fatigue Human-Automation Interaction and Safety IoT and GPS-based Vehicle Safety Systems

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
257–261
Publication date
21 January 2026
Language
EN