Exploiting the potential of eye movements analysis in the driving context

Driving is a complex and highly visual task. With the development of high-end eyetracking devices, numerous studies over the last two decades have investigated eye movements of the driver to identify deficits in visual search patterns and to derive assistive, informative, and entertainment systems. However, little is known about the visual behavior during autonomous driving, where the driver can be involved in other tasks but still has to remain attentive in order to be able to resume control of the vehicle. This work aims at exploiting the potential of eye movement analysis in the autonomous driving context. In a pilot study, we investigated whether the type of the secondary task in which the driver is involved, can be recognized solely from the eye movement parameters of the driver. Furthermore, we will discuss several applications of eye movement analysis to future autonomous driving approaches, e.g., to automatically detect whether the driver is being attentive and – when required – to guide her visual attention towards the driving task.