Functional Near Infrared Imaging for Cognitive Event Detection during Target Discrimination in Military Imagery

"With an increase in highly mobile and unconventional military threats, real-time target detection from sensor based imagery has become correspondingly more important. At the same time, computers alone have not been able to mimic the capabilities of anexpert analyst in making the visually complex decisions needed to perform high level target detection decisions. In this proposal we plan to merge human and machine capabilities to perform high-level target detections based on the monitoring of expertanalysts' cognitive performance. Using two in-house technologies, we aim to integrate functional near-infrared (fNIR) imaging with a real-time spatial-spectral target analysis and reporting system (SSTARS) designed to analyze a stream of both panchromaticand hyperspectral imagery with a number of automatic anomaly and target identification algorithms . Not only does this integration revolutionize current target detection capabilities, it will allow computers to utilize cognitively significant events tomake decisions an order of magnitude faster and more accurately than current technologies. This augmentation of both humans and computers will allow for a direct application and expansion of work already funded by DARPA's Augmented Cognition program toapply directly to a C4ISR environment. The proposed work will demonstrate, as a foundation for continued research, the feasibility of using cognitive event detection to improve false alarm