Uncertain Pointer: Situated Feedforward Visualizations for Ambiguity-Aware AR Target Selection

TL;DR

Uncertain Pointer is presented, a systematic exploration of feedforward visualizations that annotate multiple candidate targets before user confirmation, either by adding distinct visual identities to support disambiguation or by modulating visual intensity to convey system uncertainty.

Abstract

Target disambiguation is crucial in resolving input ambiguity in augmented reality (AR), especially for queries over distant objects or cluttered scenes on the go. Yet, visual feedforward techniques that support this process remain underexplored. We present Uncertain Pointer, a systematic exploration of feedforward visualizations that annotate multiple candidate targets before user confirmation, either by adding distinct visual identities (e.g., colors) to support disambiguation or by modulating visual intensity (e.g., opacity) to convey system uncertainty. First, we construct a pointer space of 25 pointers by analyzing existing placement strategies and visual signifiers used in target visualizations across 30 years of relevant literature. We then evaluate them through two online experiments (n = 60 and 40), measuring user preference, confidence, mental ease, target visibility, and identifiability across varying object distances and sparsities. Finally, from the results, we derive design recommendations in choosing different Uncertain Pointers based on AR context and disambiguation techniques.

Figures (9)

• Figure 1: Flow of information through the different phases of our systematic review, following PRISMA guidelines.
• Figure 2: Pointer space of Uncertain Pointer, defined across three dimensions: (a) Uncertainty Complexity Type, strategies for facilitating disambiguation that convey varying amounts of uncertainty information (e.g., Certain discloses no uncertainty, Identity shows the existence of uncertainty, Level reveals a graded level of uncertainty); (b) Visual Signifier, visual attributes used to represent uncertainty levels or identities (e.g., color, size, opacity, text/symbol); (c) Pointer Archetype, different forms and spatial placements of visualizations relative to the target objects (e.g., boundary, fill, internal, external).
• Figure 3: (A) Our mock-up AR video generation pipeline from Gaussian splat capture to resulting videos along with the 4 scenes used for our study with varying target distance and sparsity: (B) sparse $\times$ far, shown with a external-certain pointer, (C) dense $\times$ far, shown with internal-text-identity pointers, (D) sparse $\times$ near, shown with fill-color-identity pointers, and (E) dense $\times$ near, shown with boundary-color-level pointers, across the tower, tennis court, paintings, and books scenes, respectively.
• Figure 4: Study results for certain pointers in terms of Preference, Confidence, Mental Ease, Target Visibility, and Duration for each archetype, scene, and their combinations. The error bars represent 95% confidence intervals. All participants answered correctly on the counting task, so Error in Count is always 0 and excluded from the plot.
• Figure 5: Study results for identity pointers in terms of Preference, Confidence, Mental Ease, Target Visibility, Duration, and Error in Count for each archetype, signifierscene, and their combinations. The error bars represent 95% confidence intervals.