Designing Accessible Content Creation Support with Blind and Low Vision Creators

Abstract

Today, digital creative tools are still largely designed without disabled people in mind and thus require retrofitting to achieve basic accessibility. As more creative tools begin to incorporate generative artificial intelligence, I propose a research agenda that centers the design of human-AI co-creation experiences on disabled creators and leverages such technology for accessibility from the start. Specifically, I focus on researching ways that AI-assisted creative tools could be designed to lift the expression ceiling and reduce effort for blind and low vision creators. Starting with a formative mixed-method study that uncovers the blind and low vision community's needs for visual content creation and editing support, this dissertation explores and designs accessible content creation support for three highly desired visual creative tasks: (1) private visual content obfuscation, (2) social media video editing, and (3) aesthetic visual content authoring. Together, I believe this dissertation will contribute actionable insights to lower the barriers to expressive and efficient digital content creation for blind and low vision creators.