An Explainable Real-Time Driver Drowsiness Detection System Using Vision Transformers and Behavioral Fatigue Indicators
International Journal of Advanced Research in Science Communication and Technology · 15.04.2026 · 364
DOI: 10.48175/ijarsct-33444 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Advanced Research in Science Communication and Technology, online erschienen am 15.04.2026. Verfasst von Pisolla Rahul und 3 weiteren Autor:innen. Die Arbeit nennt 12 Literaturangaben.
Abstract (Originalsprache)
Driver drowsiness is a significant cause of road accidents, particularly during prolonged and monotonous driving. Early and reliable detection of fatigue is therefore essential for improving road safety. This paper presents an explainable real-time driver drowsiness detection system that integrates Vision Transformers (ViT) with behavioral fatigue indicators derived from facial analysis. The proposed approach is non-intrusive and operates using a standard real-time camera feed, making it suitable for practical in-vehicle deployment. Facial landmarks are extracted using MediaPipe FaceMesh from live video streams to compute interpretable indicators such as Eye Aspect Ratio (EAR), Mouth Aspect Ratio (MAR), blink frequency, and PERCLOS, which capture fatigue-related behavioral patterns. In parallel, facial image regions captured by the camera are analyzed using a Vision Transformer to perform deep learning–based alertness classification. A hybrid decision mechanism fuses heuristic indicators with transformer outputs and applies temporal smoothing to reduce false alarms. Real-time audio and visual alerts are triggered only under sustained drowsiness conditions. Experimental results on multiple public datasets and real-time camera-based evaluations demonstrate high detection accuracy with low latency. The proposed system bridges the gap between black-box deep learning models and practical, explainable driver safety solutions for intelligent transportation systems..
Zitieren
APA
Rahul, P., Afridi, M., Venkatesh, G., & Dheeraj, D. S. (2026). An Explainable Real-Time Driver Drowsiness Detection System Using Vision Transformers and Behavioral Fatigue Indicators. International Journal of Advanced Research in Science Communication and Technology, 364. https://doi.org/10.48175/ijarsct-33444
Vancouver
Rahul P, Afridi M, Venkatesh G, Dheeraj DS. An Explainable Real-Time Driver Drowsiness Detection System Using Vision Transformers and Behavioral Fatigue Indicators. International Journal of Advanced Research in Science Communication and Technology. 2026:364. doi:10.48175/ijarsct-33444
BibTeX
@article{Rahul2026,
author = {Rahul, Pisolla and Afridi, Mohammed and Venkatesh, Guguloth and Dheeraj, Dr S},
title = {An Explainable Real-Time Driver Drowsiness Detection System Using Vision Transformers and Behavioral Fatigue Indicators},
journal = {International Journal of Advanced Research in Science Communication and Technology},
year = {2026},
pages = {364},
doi = {10.48175/ijarsct-33444}
}
EUVIMED