Distraction prediction based on eye movements and driving performance
Ran Li, Jinwei Guo, Haibo Dong, Lin Zhang · 2025
Distracted driving poses a significant risk to road safety, as highlighted by the World Health Organization. This research focuses on the accurate identification and analysis of distracted driving behaviors, which can be caused by various internal and external factors. Two main types of distractions are visual and cognitive. Visual distractions involve the driver's gaze shifting away from the road, despite the visual attention of the driver being directed towards the road, can induce a state of mental meandering that diverts the cognitive resources away from the driving task. The study utilizes eye-tracking data and driving performance data to develop a comprehensive model for recognizing distracted driving, employing a CNN-LSTM network. The findings reveal that distraction significantly affects both eye movements and driving performance, and the proposed model demonstrates high accuracy in identifying distracted driving states.