DIYA by Vision-Aid

People, evidence and craft.

DIYA is shaped by visually impaired contributors and volunteer subject-matter experts, including Dr. Sarika Gopalakrishnan, Dr. Amy, Dr. Aparna and Dr. Arathy. Their feedback informs the vision patterns, rehabilitation examples, comfortable pacing and practical accessibility comparisons.

Clinical learning references

Clinical examples draw on National Eye Institute information and specialist feedback. This project is an educational illustration, not a clinical assessment tool.

• Low vision and rehabilitation
• Macular degeneration and central vision
• Glaucoma
• Retinitis pigmentosa
• Diabetic retinopathy
• Cataract

The colour example uses the deutan tables from Machado, Oliveira and Fernandes (2009), A Physiologically-based Model for Simulation of Color Vision Deficiency. The renderer interpolates the published table in linear-light RGB. It is an illustration, not a colour-vision test.

Detailed clinical and rendering notes

3D environment

The coherent street is adapted from Amazon Lumberyard Bistro, published through NVIDIA's Open Research Content Archive. The original environment is by Amazon Lumberyard (2017) and is licensed under Creative Commons Attribution 4.0 International.

DIYA's changes include an exterior-only selection, metric positioning, removal of shop lettering, advertisements and product or portrait imagery, a flush crossing ramp in the original pavement, detail reduction and lighter embedded textures. The crossing equipment, vision effects, hands-first interaction and educational story are part of the DIYA application. The scenery is licensed third-party work; no endorsement by Amazon or NVIDIA is implied.

Full attribution and modifications · Original license · Bundled asset record

Vehicles

The unbranded sedan is a 3D asset from the Innerscene public-domain library.

• Graphite four-door sedan

See the library terms, asset credits and source manifest.

Software and narration

The application uses Meta's Immersive Web SDK, Three.js, TypeScript and Vite. Dependency notices remain with the source package. AI coding assistance supported development. Narration is synthetic: Aoede and Orus voices are bundled locally so a live voice service is not required during the experience.

Narration text, voice and file record

The separate background film

The optional V9 excerpt combines project education with reused Vision-Aid program imagery. Other V9 setting illustrations use generated footage. That earlier film is kept separate from proof of the interactive app and is not presented as headset capture, a real street recording or a 360-degree environment.