Researchers Build A11yShape To Enable 3D Modeling
Researchers from the University of Washington, Purdue, MIT, HKUST, Stanford, Michigan, Texas at Dallas, and NVIDIA developed A11yShape to help blind and partially-sighted developers create, edit, and verify 3D models. Presented at ASSETS '25, A11yShape uses OpenAI's GPT-4o to convert OpenSCAD-generated images into tokenized textual descriptions, enabling iterative, screen-reader-compatible 3D modeling workflows for fabrication.
Key Points
- 1Implements A11yShape using GPT-4o to convert OpenSCAD models into accessible textual descriptions
- 2Addresses accessibility gap enabling blind developers to create, edit, and verify 3D designs independently
- 3Provides iterative, screen-reader-friendly workflow compatible with existing assistive technologies for fabrication pipelines
Scoring Rationale
Practical accessibility advance with conference peer-reviewed credibility; scope limited to assistive 3D modeling and specific tooling.
Sources
Public references used for this report.
Practice with real Ad Tech data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Ad Tech problems

