Guiding Gaze: Comparing Cues for Visual Search

Kelley, Brendan; Wickens, Chris; Clegg, Benjamin A.; Warden, Amelia C.; Ortega, Francisco R. · 2024 · Unknown

DOI: 10.1109/vrw62533.2024.00323

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the effectiveness of augmented reality (AR) visual cues in facilitating visual search tasks, a critical component in time-sensitive scenarios such as military operations, crisis response, and search and rescue. The research addresses the need to optimize AR interface designs to reduce search time and increase accuracy, thereby improving operational efficiency and safety. Specifically, the authors compare three distinct cueing mechanisms—3D Arrows, 2D Wedges, and Gaze Lines—against a baseline condition with no visual cues to determine which design best supports user performance. The experimental design utilized a Magic Leap 2 AR head-mounted display to render a simulated environment resembling the side of a building with a 4x8 grid of 32 windows. Participants, consisting of 14 individuals with a mean age of 22.79, performed 256 searches in total, completing 64 trials for each of the four conditions (No Cue, 3D Arrow, 2D Wedge, and Gaze Line). The task involved locating a specific target wearing a military vest and helmet among 19 distractors and 12 empty windows. The order of conditions was counterbalanced using Latin squares to mitigate learning effects. Data collected included search response time and accuracy rates. Results from a one-way ANOVA indicated statistically significant differences across all conditions for both response time and accuracy. All cueing conditions significantly outperformed the No Cue baseline, which had an average search time of 6.16 seconds and an accuracy of 87.16%. The Gaze Line cue yielded the best performance, reducing search time to 1.35 seconds and achieving 99.11% accuracy. The 2D Wedge cue followed closely with a search time of 1.88 seconds and 97.43% accuracy. The 3D Arrow cue was the least effective of the three, resulting in a search time of 2.87 seconds, though it still improved accuracy to approximately 94% (inferred from the 7% increase mentioned in the discussion). Statistical comparisons confirmed that Gaze Lines were significantly faster than 2D Wedges, which were in turn significantly faster than 3D Arrows. The authors conclude that while any visual cue improves performance over no cue, the design’s ability to communicate spatial information is crucial. The superior performance of Gaze Lines is attributed to their intuitive nature and their ability to explicitly communicate both direction and target position. In contrast, 3D Arrows only indicate direction, requiring users to extrapolate position, which introduces ambiguity and slows performance. The study suggests that future AR interface designs for visual search should prioritize cues that provide clear, intuitive spatial guidance, such as Gaze Lines, to maximize efficiency in high-stakes environments. Limitations noted include the use of a forward-facing search field and accurate cues, suggesting further research is needed for 360-degree fields and erroneous cueing scenarios.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-08 (2 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-09; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• visual search