Effects of Clutter and AI Cueing in Visual Search Tasks

Visual clutter impairs visual search by increasing cognitive load and reducing target discriminability (Moacdieh & Sarter, 2015). Clutter can arise from multiple sources: display density (number and density of elements), display layout (spatial organization, feature variability), and target-background similarity (shared visual features reducing target salience). These factors are particularly relevant in safety-critical domains such as maritime search and rescue, where operators must detect small targets in cluttered aerial imagery under perceptual constraints (Frost, 1996). AI-based detection systems can mitigate perceptual challenges by augmenting human attention (Mosier & Skitka, 1996). However, the benefits of AI support are not guaranteed. Operators may inappropriately use the system (Parasuraman & Riley, 1997), leading to miscalibrated trust that manifests as compliance (agreement with target-present recommendations) and reliance (agreement with target-absent recommendations; Meyer, 2001, 2004). Achieving optimal joint performance requires operators to avoid both overtrust and distrust (Lee & See, 2004). Cueing benefits depend on display conditions, with greater improvements under perceptually demanding conditions (Maltz & Shinar, 2003). A prior study (own work, in preparation) compared three cueing approaches (textual alerts, bounding boxes, and contextual blur) across color and monochrome displays. Under monochrome conditions, direct cues (bounding boxes, contextual blur) yielded higher detection rates, better discrimination, and greater compliance than indirect cues (textual alerts). Planned contrasts comparing bounding boxes (additive cueing, preserving scene context) with contextual blur (subtractive cueing, suppressing surrounding information) revealed a display-dependent dissociation: Box produced higher accuracy, sensitivity, and a more conservative response criterion than Blur in monochrome, while Blur produced higher compliance. Under color conditions, these differences vanished entirely. The present study investigates how visual clutter moderates the effectiveness of bounding boxes (Box) and contextual blur (Blur) across color and monochrome display conditions. Box provides a spatial anchor within the cluttered scene while preserving surrounding context, whereas Blur suppresses distracting background elements. We examine how display type, cueing method, and visual complexity jointly affect performance, signal detection measures, trust behavior, and subjective experience. Of particular interest is the interplay between cueing method and visual complexity under monochrome conditions.