Thesis
Improving Anomaly Detection through Identification of Physiological Signatures of Unconscious Awareness
(2016)
Abstract
Missed anomalies have the potential to cause detrimental effects in the Intelligence, Surveillance, and Reconnaissance (ISR) domain. One possible cause of these missed anomalies is that cognitive processing may not reach conscious awareness and may only be perceived by the unconscious mind. Identification of correlates of these unconscious processes could provide an insight into potential missed targets. The present study explored missed anomalies in a visual search task and the possibility of unconscious awareness. Eye metrics were recorded and a "Detection Threshold Model" was created and validated with a nominal logistic regression model, in order to characterize the search patterns and eye metrics of detection, non-detection, and possible unconscious detection. Results indicated that eye metrics of fixation count, fixation duration, mean saccade length, and backtrack rate predicted detections and non-detections with an overall accuracy of about 90%. Additionally, gaze plots of possible unconscious detections revealed signature search patterns of detection.
Keywords
- Eye metrics
Disciplines
Publication Date
Spring April 20, 2016
Degree
Master of Science in Biomedical Engineering (MSBME)
Department
Biomedical, Industrial & Human Factors Engineering
Citation Information
Alyssa Marie Piasecki. "Improving Anomaly Detection through Identification of Physiological Signatures of Unconscious Awareness" (2016) Available at: http://works.bepress.com/nasser_kashou/53/