Answers are grounded in this paper and cite their sources.
Artificial intelligence (AI)-based detection systems can enhance visual search in safety-critical domains, yet how they present recommendations may determine their effectiveness under display-quality constraints. We compared three cueing methods: an indirect cue signaling only presence (textual alert) and two direct cues marking the target’s location through either signal enhancement (bounding box) or noise reduction (contextual blur), across color and monochrome displays. Participants (N = 122) searched maritime aerial imagery in a fixed-order block structure (unaided, aided, unaided). Direct cues improved detection only under monochrome conditions, whereas the indirect cue lowered joint human–AI accuracy below that of the AI alone. Bounding boxes yielded the highest detection, whereas contextual blur reduced response times and increased compliance. Temporal analyses revealed that aided performance diverged from retained skill, with blur the most durable after withdrawal despite stronger signs of automation dependence. Cueing should be matched to display type and operational context, and judged by the skill that persists after support.
No takes yet. Share an insight, caveat, or question.
Gruner et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: