What is “equivalent” text alternatives for artistic and cultural visual content
summary
Library of Congress runs a 2024 research program testing image description workflows with six deep-interview participants across 17–18 images and assistive tech (screen readers, magnifiers, braille). Early results show users want short alt + rich longdesc (33% alt only, 37% longdesc, 28% inline text, 2% external file) and high demand for facts: 92% want a generic summary, 91% text content, 90% emotive/motive clues, 88% context, while only 40% request visual quality. Guidance favors explicit identity when known and using OCR/AI to surface queryable details.
AI-written summary · about
chapters (13)
- 0:00Research introduction
- 2:49Study scope and objectives
- 5:11Defining image descriptions
- 6:57Categories of description
- 8:21Practical categorization examples
- 13:57Image case studies
- 27:35Floor plan study insights
- 31:49Qualitative study results
- 33:06Best practices for controls
- 35:21Detail and complexity findings
- 38:30Guidance and key takeaways
- 46:06Future research and AI potential
- 48:04Q&A and concluding remarks
related talks
20mRich Representations of Visual Content for Screen Reader Users
01h 00mAI and alt text
16mThe positive alt-itude! An inclusion tool for your accessibility
16mCaption Crawler: Enabling Reusable Alternative Text Descriptions using Reverse Image Search
17mPositive alt-itude! An Inclusion Tool for Your Accessibility
22mFAULT TEXT: The Adversities in Automated Alternative Text with Henri Helvetica
56mBeyond Colors and Captions: How to Provide More Inclusive Accessibility
55mBridging the Digital Divide: Empowering Accessibility Through AI Innovations