Evaluating Driver Features for Cognitive Distraction Detection and Validation in Manual and Level 2 Automated Driving

Abstract

Objective This study aimed to investigate the impacts of feature selection on driver cognitive distraction (CD) detection and validation in real-world nonautomated and Level 2 automated driving scenarios. Background Real-time driver state monitoring is critical to promote road user safety. Method Twenty-four participants were recruited to drive a Tesla Model S in manual and Autopilot modes on the highway while engaging in the N-back task. In each driving mode, CD was classified by the random forest algorithm built on three “hand-crafted” glance features (i.e., percent road center [PRC], the standard deviation of gaze pitch, and yaw angles), or through a large number of features that were transformed from the output of a driver monitoring system (DMS) and other sensing systems. Results In manual driving, the small set of glance features was as effective as the large set of machine-generated features in terms of classification accuracy. Whereas in Level 2 automated driving, both glance and vehicle features were less sensitive to CD. The glance features also revealed that the misclassified driver state was the result of the dynamic fluctuations and individual differences of cognitive loads under CD. Conclusion Glance metrics are critical for the detection and validation of CD in on-road driving. Applications The paper suggests the practical value of human factors domain knowledge in feature selection and ground truth validation for the development of driver monitoring technologies.

Investigating the role of flight phase and task difficulty on low-time pilot performance, gaze dynamics and subjective situation awareness during simulated flight

“…This is a critical aspect of gaze behaviour to examine as a lack of scanning during phase(s) of flight that require continuous monitoring of flight parameters (i.e., landing) and aircraft state via gauges inside the cockpit can result in disastrous consequences should an unexpected/hazardous event occur. This phenomenon was similarly reported in driving simulation studies that found that the higher proportion of dwell time allocated to the center of the road and a reduction in gaze dispersion were associated with driver distraction, reduced hazardous event detection, and increased driver cognitive load ( 41 ; 58 ; 60 ; 61 ). The increased cognitive tunneling findings seemingly contradict the increase in final approach SGE and GTE during difficult trials.…”

Temporal Dashboard Gaze Variance (TDGV) Changes for Measuring Cognitive Distraction While Driving

“…This performance is on par with current state-of-the-art machine learning methods, such as from Yang et al 2020 [ 22 ], who developed three models for detecting cognitive distraction with accuracies around 80%. Similarly, the AIDE EU project states that for the purpose of informing other in-vehicle systems, a distraction detection accuracy of 70% is sufficient and 85% is good.…”

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

“…Abosaq et al [ 65 ] proposed a customized CNN model ( Figure 5 ) to recognize normal and abnormal driver actions (including driver smoking, driver eating, driver drinking, driver calling, and driver normal) from driver videos, and achieved 95% accuracy on the prepared testing dataset. Yang et al [ 66 ] investigated the impacts of feature selection on driver cognitive distraction detection and validation in real-world non-automated and Level 2 automated driving scenarios. A Mobileye sensor recorded vehicle performance while two Logitech webcams and a forward-facing camera collected video data of 24 drivers (12 males and 12 females with ages 22–68) and roadway.…”