What to Distinguish and How? Opportunities and Challenges of Augmenting Multiple, Cluttered Objects in Complex Scenes for People with Low Vision

Jul 16, 2026

Upload by Sage

People with low vision (PLV) struggle to perceive complex scenes such as busy kitchens and crowded streets, where many objects, visual clutter, and dynamic elements compete for attention. Prior AR approaches largely enhanced low-level features or augmented single task-relevant objects, leaving multiobject augmentation in complex scenes insufficiently studied. This work presents SceneGlance, a wearable AR system that recognizes important objects and distinguishes them by perceived importance. Controlled and think-aloud studies show improved attention and hierarchical scanning, with a recall tradeoff and practical challenges like adjacent augmentation blending.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages24

Reading Time1 h 0 min

Introduction

Motivation and background

Prior work in AR for low vision

SceneGlance approach

Evaluation studies

Controlled lab study

Free-form think-aloud study

Findings and design implications

Attention shift and recall tradeoff

Visual blending challenges in cluttered scenes

What to Distinguish and How? Opportunities and Challenges of Augmenting Multiple, Cluttered Objects in Complex Scenes for People with Low Vision arXiv :2607 . 10902v1 [ cs .HC] 12 Jul 2026 Yuheng Wu [yuheng.wu@wisc.edu](yuheng.wu@wisc.edu)[ ](yuheng.wu@wisc.edu)University of Wisconsin-Madison Madison, WI, USA Jia Li [jia.li@utdallas.edu](jia.li@utdallas.edu)[ ](jia.li@utdallas.edu)University of Texas at Dallas Richardson, TX, USA Weibing Wang [wwang652@wisc.edu](wwang652@wisc.edu)[ ](wwang652@wisc.edu)University of Wisconsin-Madison Madison, WI, USA Ruijia Chen [ruijia.chen@wisc.edu](ruijia.chen@wisc.edu)[ ](ruijia.chen@wisc.edu)University of Wisconsin-Madison Madison, WI, USA Kexin Zhang [kzhang284@wisc.edu](kzhang284@wisc.edu)[ ](kzhang284@wisc.edu)University of Wisconsin-Madison Madison, WI, USA Sanbrita Mondal [smondal4@wisc.edu](smondal4@wisc.edu)[ ](smondal4@wisc.edu)University of Wisconsin-Madison Madison, WI, USA Jaewook Lee [jaewook4@cs.washington.edu](jaewook4@cs.washington.edu)[ ](jaewook4@cs.washington.edu)University of Washington Seattle, WA, USA Meng Fong Lio [mengfong@wisc.edu](mengfong@wisc.edu)[ ](mengfong@wisc.edu)University of Wisconsin-Madison Madison, WI, USA Jon E. Froehlich [jonf@cs.uw.edu](jonf@cs.uw.edu)[ ](jonf@cs.uw.edu)University of Washington Seattle, WA, USA Yapeng Tian [yapeng.tian@utdallas.edu](yapeng.tian@utdallas.edu)[ ](yapeng.tian@utdallas.edu)University of Texas at Dallas Richardson, TX, USA Yuhang Zhao [yuhang.zhao@cs.wisc.edu](yuhang.zhao@cs.wisc.edu)[ ](yuhang.zhao@cs.wisc.edu)University of Wisconsin-Madison Madison, WI, USA A B Figure 1: We explore the opportunities and design challenges of augmenting and distinguishing multiple objects in complex scenes to support scene perception for people with low vision, using SceneGlance as a technology probe. (A) SceneGlance distinguishes multiple objects in a complex kitchen, with primary-important objects (glasses, knives) overlaid in solid yellow and secondary objects (bowls, spoons, ladles) in a blue outline. (B) SceneGlance distinguishes multiple objects on a crowded street, with primary-important objects (curb, sewer drain, pedestrian signal, bicycles) in a green outline with floating green icon labels, and secondary objects (sidewalk, pedestrian) in a blue outline only. This work is licensed under a Creative Commons Attribution 4 .0 International License. ASSETS’26, Vila Nova de Gaia, Portugal © 2026 Copyright held by the owner/author(s) . ACM ISBN 979-8-4007-2521-0/2026/10 [https://doi.org/10.1145/3797867.3828974](https://doi.org/10.1145/3797867.3828974) Abstract People with low vision (PLV) struggle to perceive complex scenes like busy kitchens and crowded streets, which contain many objects, visual clutter, and dynamic elements. Prior AR systems for low vision either enhance low-level visual features or augment taskrelevant objects for single tasks in simple settings, leaving multiobject augmentation in complex scenes underexplored. Informed by a formative study characterizing important objects and their perceived importance for PLV, we built SceneGlance, a wearable AR system that recognizes important objects and visually distinguishes them by importance level. Through a controlled lab study with 12 PLV in a mock-up kitchen scene and a free-form think-aloud study with 13 PLV navigating an outdoor route, we found that AR distinction on object importance shifted PLV’s attention toward objects of higher importance, and supported perception strategies such as building mental snapshots from augmentation distribution and hierarchical scanning by importance. However, this attention shift came with a tradeoff, as augmenting many objects reduced overall scene recall. The studies also surfaced challenges posed by AR augmentations in complex scenes, such as adjacent augmentations blending or interfering with each other, yielding design implications for more practical AR vision enhan

FAQs

What problem does the paper address for people with low vision?

People with low vision face difficulty perceiving complex scenes that contain many objects, visual clutter, and dynamic elements, such as kitchens and crowded streets.

How does SceneGlance distinguish objects in complex scenes?

SceneGlance recognizes important objects and visually distinguishes them according to importance level, overlaying primary-important objects more prominently and secondary objects less prominently.

What tradeoffs and challenges did the studies reveal?

Object-importance distinction shifted attention toward more important objects and supported hierarchical scanning strategies, but augmenting many objects reduced overall scene recall. The work also found challenges such as adjacent augmentations blending or interfering with each other.

Jul 16, 2026

Upload by Sage