Machine Detection of Operationally Significant Cognitive Events for C4ISR

摘要

A machine capable of detecting cognitively significant events in its user could prevent potential disaster by signaling to commanders that a soldier is under high stress. This project seeks to establish that these cognitive events can be captured in an autonomous fashion through the use of an eye- tracking system. The experiment in this study requires subjects to find a particular person hidden in a sequence of complex images that contain crowded scenes of different people performing different activities. Project tasks included creating the test stimulus, running test subjects, and analyzing the captured data. This analysis indicates that a pupil dilation increase during a period of prolonged fixation occurs when the test subject finds the target person in the stimulus. Additional testing is necessary to validate this finding in a more realistic setting, but this study represents a preliminary step in developing a machine capable of autonomously detecting cognitive events.