Gaze Movement’s Entropy Analysis to Detect Workload Levels

The gaze movement from a driver represents specific skills related to safe driving. Driving maneuvering evaluation is to know driving fitness. We found that gaze movement entropy is highly sensitive to visual behavior demands while driving a vehicle and the workload level. Entropy measures were more sensitive, more robust, and easier to calculate than gaze established measures. The gaze movement measures were collected using a driving simulator, using five different simulate routes and tree different workload scenarios; one route of familiarization was used to create a baseline. Because the workload became more difficult, drivers looked more at the central part of the road for more extended periods, the gaze movement entropy values decreased when the workload was increased, and the results show differences between two levels of workload. Also, the entropy result is compared against the classical analysis of the spatial distribution of gaze.