Automatic Alt Text Generator

MDP

Problem

Websites need to meet accessibility standards and remove barriers for individuals using screen readers. Developers often spend significant time creating descriptive alt text for images, which is labor-intensive and prone to inconsistency.

Audience

The tool is designed for web developers and content creators who manage digital assets on university platforms.

Outcome/Impact

Solution: as part of a Multidisciplinary Design Program (MDP) project, a team of University of Michigan students developed three tools leveraging generative AI to automatically create alternative text, commonly called Alt-Text, to create image captions. These tools were implemented as a WordPress plugin, Drupal module, and Google Docs and Slides add-on, providing compatibility across widely used platforms. The project emphasized accessibility and innovation, showcasing student expertise and collaboration.

  • Improved Accessibility: Increases the usability of websites for individuals relying on screen readers.

  • Efficiency Gains: Automates the creation of alt text, saving time and resources for developers.

  • Customization: Allows users to modify system prompts and regenerate alt text for better alignment with specific requirements.

  • Scalability: Offers multi-upload capabilities for large-scale image management.

  • Compliance: Helps ensure adherence to WCAG guidelines and accessibility mandates.

Precursor to Wider Applications: While initially targeted for accessibility improvements, the tool demonstrates a scalable approach that could be applied to broader digital asset management challenges across industries.

Project Team:

  • Audrey June Smith

  • Joshua Xie

  • Karthik Tilak

  • Liyufei Meng

  • Kushal Kamal Patel

  • Yushu He