Demo: Evaluating Attention Vulnerabilities to Distraction with an AR Trail Making Test (AR-TMT)

We present AR-TMT, an AR adaptation of the Trail Making Test designed to evaluate how different types of distractors affect user attention in the AR environment. Built on the Magic Leap 2, AR-TMT captures gaze behavior as users perform a cognitively demanding visual search task under two distraction conditions: bottom-up (visually salient) and top-down (semantically misleading). The system visualizes gaze patterns and performance, revealing distinct attentional responses to the two distraction types. This demo could give insights on cognitive security in XR by offering a empirical findings of visual attention vulnerabilities with regard to distraction types and informing the design of distraction-aware, cognitively resilient AR systems.

Duke Scholars

Author Maria Gorlatova Electrical and Computer Engineering