Emerging Practices for Large Multimodal Model (LMM) Assistance for People with Visual Impairments: Implications for Design

TL;DR

This study explores how large multimodal models assist visually impaired users, revealing their adaptive use and cognitive extension, and discusses design implications for more effective assistive AI tools.

Contribution

It provides empirical insights into user interactions with LMM-based assistive tools and suggests design directions for goal-oriented, real-time AI assistance.

Findings

Users adapt tools for complex social and spatial tasks.

Tools extend users' cognition by distributing visual understanding.

Users accept current limitations and utilize broad capabilities.

Abstract

People with visual impairments perceive their environment non-visually and often use AI-powered assistive tools to obtain textual descriptions of visual information. Recent large vision-language model-based AI-powered tools like Be My AI are more capable of understanding users' inquiries in natural language and describing the scene in audible text; however, the extent to which these tools are useful to visually impaired users is currently understudied. This paper aims to fill this gap. Our study with 14 visually impaired users reveals that they are adapting these tools organically -- not only can these tools facilitate complex interactions in household, spatial, and social contexts, but they also act as an extension of users' cognition, as if the cognition were distributed in the visual information. We also found that although the tools are currently not goal-oriented, users accommodate this limitation and embrace the tools' capabilities for broader use. These findings enable us to envision design implications for creating more goal-oriented, real-time processing, and reliable AI-powered assistive technology.