Enhancing Video Communication Experience for Low Vision Users
Andreas Sackl, Raimund Schatz, Bruno Gardlo, Manfred Tscheligi · 2021 · Proceedings of the 18th International Web for All Conference (W4A) · doi:10.1145/3430263.3452419
Summary
This paper investigates whether real-time video enhancement techniques can improve the video chat experience for people with low vision. The researchers extended an existing custom video player with video chat capabilities, allowing users to individually adjust the visual appearance of their chat partner's video feed using parameters including contrast adaptation, brightness, saturation, edge detection and enhancement, and grayscale modes. Uniquely, the system also supported a tangible interface (Palette Gear) with physical rotary buttons for adjusting filters, alongside standard mouse and keyboard controls. Seven participants with diverse low vision conditions — including high-grade myopia, cataract, retinal detachment, albinism, tunnel vision, glaucoma, and macular degeneration — completed two tasks: a naturalistic 10-minute interview conversation with a study assistant, and a facial emotion recognition task using Ekman's six basic emotions (anger, disgust, fear, happiness, sadness, surprise). The study was motivated by the COVID-19 shift to remote communication and the observation that mainstream video chat tools like Zoom are inaccessible for low vision users due to small, low-contrast interfaces and no options to enhance the video content itself. Previous research had focused on detecting conversational cues or recognizing faces for blind users, but not on optimizing the actual video presentation for people with residual vision.
Key findings
All seven participants chose different enhancement configurations tailored to their individual visual impairments, confirming that a one-size-fits-all approach is inadequate. Brightness was adjusted by all participants (ranging from +3% to +57%), while contrast (+3% to +17%) and saturation (-17% to +10%) were modified by most. Five of seven activated edge detection/enhancement (+3% to +70%), and two activated grayscale mode. For the interview task, participants reported that enhancements made the experience feel more personal and realistic — one participant described it as feeling like a direct in-person conversation. For facial emotion recognition, enhanced video led to improved identification of anger (5 to 6 correct), fear (1 to 2), and sadness (4 to 6), while disgust, joy, and surprise remained the same. Participants noted that without enhancement, emotion recognition was exhausting and took longer. Several participants emphasized that existing video chat tools are too inhomogeneous with small fonts and low contrast, and that common tasks like answering or initiating a call are overly complex for low vision users. Multi-participant video calls were flagged as even more challenging than one-to-one conversations.
Relevance
This study addresses an increasingly critical accessibility gap: as video communication becomes essential for work and social life, people with low vision are being excluded by mainstream tools that offer no meaningful visual customization of the video feed itself. The key insight is that enhancing the transmitted video content — not just the application UI — can significantly improve both the communication experience and the ability to perceive facial emotions, which are essential social cues often lost for low vision users. The finding that every participant chose different enhancement settings reinforces the need for personalizable accessibility features rather than fixed accommodations. For practitioners building video communication tools, the research suggests that contrast, brightness, edge enhancement, and grayscale controls should be exposed as user-configurable options. The tangible interface approach also demonstrates the value of multimodal interaction for making real-time adjustments accessible. Limitations include the small sample size (N=7), ideal lab conditions with high-quality equipment, and restriction to bilateral (two-person) video calls.
Tags: low vision · video communication · visual enhancement · user experience · assistive technology · telepresence · facial emotion recognition
Standards referenced: WCAG 2.1