Visual Attention During Simulated Autonomous Driving in the US and Japan

Abstract

To explore cultural differences in driver behavior for the purposes of vehicle automation, we used eye tracking to measure fixation patterns of Japanese and US participants (N = 98) viewing video simulations of automated driving through San Francisco and Osaka. After each drive, we asked participants questions about objects and events from the video. Japanese participants showed higher fixation counts and durations than US participants for salient foreground objects in the traffic scene, and answered questions about those objects more correctly. US participants showed higher fixation counts than Japanese participants on visually prominent background features, and gave more correct answers about those. Consistency between fixation count and fixation duration on an object or feature, and recall of that object or feature, suggests that situation awareness during simulated automated driving proceeds from patterns of visual fixation on elements of the traffic scene.