Answers are grounded in this paper and cite their sources.
Driving is a daily activity that can lead to serious accidents when a driver’s alertness decreases. Intelligent driver monitoring systems with warning and control circuits can detect reduced alertness, issue timely alerts, and, if necessary, automatically control the vehicle to prevent collisions. Two significant factors affecting alertness are stress and drowsiness. Stress can accumulate rapidly and impair decision-making in critical situations, while drowsiness increases the risk of unsafe behaviors such as drifting off the road, speeding, or ignoring hazard lights. This study proposes a combined transformer-feedback network to detect both stress and drowsiness using electrocardiogram (ECG) signals, which reflect the driver’s physiological state and cannot be intentionally controlled. The transformer network extracts high-level temporal features, while the LSTM classifier processes sequential patterns for accurate categorization. By leveraging the strengths of the Vision Transformer (ViT) for time-series representation and LSTM for classification, the proposed method improves the detection accuracy of both factors compared with recent studies that typically focus on only one aspect of driver behavior. Overall, this unified framework simplifies the implementation of two mechanisms by using a single model across two applications, offering a practical and efficient solution for intelligent driver-monitoring and alert systems integrated into vehicles.
No takes yet. Share an insight, caveat, or question.
Al-Khuwaylidee et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: