Detecting drivers' cognitive load from saccadic intrusion. (April 2018)

• Record Type: Journal Article
• Authors: Biswas, Pradipta; Prabhakar, Gowdham
• Is Part Of: Transportation research. Volume 54(2018)
• Page Start: 63
• Page End: 78
• Publication Date: 2018-04
• Subjects: Eye gaze tracker -- Automotive -- Saccadic intrusion -- Cognitive load -- Hazard perception -- Human machine interaction; Automobile drivers -- Psychology -- Periodicals; Automobile driving -- Psychological aspects -- Periodicals; Transportation -- Psychological aspects -- Periodicals; 629.283019
• DOI: 10.1016/j.trf.2018.01.017
• Languages: English
• ISSNs: 1369-8478

Abstract

Graphical abstract: Saccadic intrusions are conjugate, horizontal saccadic movements which tend to be three to four times larger than the physiological microsaccades and take the form of an initial fast eye movement away from the desired eye position, followed, after a variable duration, by either a return saccade or a drift. A certain type saccadic intrusion having velocities more than 3°/sec can be used to detect change in cognitive load as well as developing road hazard. Highlights: Saccadicintrusions(SI) are conjugate, horizontal micro-saccadic movements. Velocities of saccadic intrusion (SI) vary across participants. Low velocity SI kept on occurring even in situations demanding less cognitive load. With increase in demand of cognitive load, the average velocity of SI increases. Users' instantaneous perception of developing road hazard accompanies high velocity SI. SI having velocities more than 3°/sec can be used to detect change in cognitive load. Abstract: In this paper, we have explored use of Saccadic Intrusion to detect drivers' cognitive load and instantaneous perception of developing road hazards. Saccadic Intrusion is a type of eye gaze movement which was earlier found to be related to change in cognitive load. We have developed an algorithm to detect saccadic intrusion from a commercially available low-cost eye gaze tracker and conducted four user studies involving a driving simulator and cognitive and hazard perception tests. Our results show that average velocities of saccadic intrusion increases with increase in cognitive load and recording saccadic intrusion and eye blinks for 6 s duration can predict drivers' instantaneous perception of developing road hazards.