Application of Gaze-Based Adaptive Notifications to Support Image Analysts

An application that emulates the working environment of an image analyst has been developed. An eye-tracker is used to observe the viewing behavior and based on this viewing behavior and image analyst’s computer keyboard interaction, the application tries to optimize the presentation of notifications (namely, adaptive notifications). Within the study, three notification modes were explored: adaptive notifications, non-adaptive notifications and no notifications. The goal of the study was to explore whether adaptive notifications result in higher task performance, situational awareness, and sense of trust, and a lower sense of irritation. A total of fifteen participants participated. They observed two simultaneous video streams and reported mission-relevant information. Each participant performed three scenarios with a different notification mode in each scenario. Task performance was logged in the background by the application. Results indicate that task performance was higher when using the adaptive notification mode compared to the other modes. More specifically, the mode with adapted notifications resulted in fewer false alarms and more correct rejections compared to other modes. There were no significant differences in situational awareness, trust and irritation between the different notification modes. Adaptive notifications based on gaze input can increase the image analyst’s task performance.