Classification of Driver Distraction Risk Levels: Based on Driver’s Gaze and Secondary Driving Tasks

Abstract

Driver distraction is one of the significant causes of traffic accidents. To improve the accuracy of accident occurrence prediction under driver distraction and to provide graded warnings, it is necessary to classify the level of driver distraction. Based on naturalistic driving study data, distraction risk levels are classified using the driver’s gaze and secondary driving tasks in this paper. The classification results are then combined with road environment factors for accident occurrence prediction. Two ways are suggested to classify driver distraction risk levels in this study: one is to divide it into three levels based on the driver’s gaze and the AttenD algorithm, and the other is to divide it into six levels based on secondary driving tasks and odds ratio. Random Forest, AdaBoost, and XGBoost are used to predict accident occurrence by combining the classification results, driver characteristics, and road environment factors. The results show that the classification of distraction risk levels helps improve the model prediction accuracy. The classification based on the driver’s gaze is better than that based on secondary driving tasks. The classification method can be applied to accident risk prediction and further driving risk warning.

Comprehensive Assessment of Artificial Intelligence Tools for Driver Monitoring and Analyzing Safety Critical Events in Vehicles

“…The RiskNet was trained on a simulated collision dataset (58,904 safe and 7788 unsafe frames) and tested on real-world collision dataset (3604 safe and 1008 unsafe frames) with an accuracy of 91.8% and F1-score of 0.92. In another study, Zheng et al [ 83 ] used naturalistic driving data from The 100-Car Naturalistic Driving Study to classify distraction risk levels via driver’s gaze or secondary driving tasks. They combined distraction risk levels, road environment factors, and driver characteristics to predict influencing factors on accident occurrence via random forest, AdaBoost, and XGBoost.…”

Analysis of factors affecting crash under risk scenarios based on driver homogenous clustering

“…crashing with roadblocks, animals, etc; Pre-incident maneuver describes whether the driver has accelerated or steered prior to the risk event; Driver reaction refers to the driver's crash avoidance operation for the risk scenario, which contains explicitly four types of braking, steering, no reaction, and acceleration; Distraction state refers to the severity of a driver's distraction when a risk scenario occurs, and distractions can be classified into four risk levels based on the impact of various distractions on the outcome of the risk event [20]; Multiple…”

“…Beyond the aforementioned issues, there are certain shortcomings in the quantification methods applied to assess the effects of influencing factors in existing studies. These studies predominantly rely on regression techniques, with logistic regression being a prime example [20,21]. However, it's crucial to recognize that traditional regression methods have certain limitations when applied to the domain of traffic safety analysis.…”