Aided target recognition visual design impacts on cognition in simulated augmented reality

Abstract

Aided target recognition (AiTR) systems, implemented in head-mounted and in-vehicle augmented reality (AR) displays, can enhance human performance in military operations. However, the visual appearance and delivery of AiTR may impact other important critical aspects of human performance like decision making and situational awareness (SA). Previous research suggests salient visual AR cueing, such as found in Computer-Aided Detection diagnostic systems, orient attention strongly toward cued targets leading to missed uncued targets, an effect which may be lessened by providing analog information about classification uncertainty and using less visually salient cueing techniques, such as soft highlighting. The objective of this research was to quantify the human performance impacts of two different types of AR AiTR visualizations in a simulated virtual reality defensive security task. Participants engaged in a visual camouflage discrimination task and a secondary SA Task in which participants observed and reported a peripheral human target. Critically, we manipulated the type of AiTR visualization used: 1) a traditional salient bounding box, 2) a softly glowing soft highlight, and 3) a baseline no-AiTR condition. Results revealed minimal impacts of the visual appearance of AiTR on target acquisition, target categorization, and SA but an observable reduction in user experience associated with soft highlight AiTR. Future research is needed to explore novel AiTR designs that effectively cue attention, intuitively and interpretably visualize uncertainty, and deliver acceptable user experience.