Skip to content
All work

Vision Aid

iOSDissertation
Vision Aid — app screenshots

Overview

Vision Aid is an assistive iOS app built for blind users. It helps them understand and navigate their surroundings by applying machine learning to the device’s sensor input and turning the results into sound. It was my first app, built as my dissertation project.

My role

I designed and developed the app on my own, from the research and accessibility design through to the implementation.

Hardest part

Designing an interface nobody will look at. Every convention I knew, visual hierarchy, tappable targets, on-screen feedback, assumes a user who can see the screen, and none of them apply here.

I built the app output-first: results are conveyed through sound rather than shown, controls work without precise aim, and the whole app is navigable by VoiceOver, so the screen is never the thing the user has to fight.

Architecture & stack

Swift and UIKit, using CoreML and Vision over the device’s sensors and camera input, with audio feedback and full VoiceOver support as the primary interface.

Achievements

  • Built as my Computer Science dissertation at the University of Sheffield
  • My first complete iOS app, and the one that started the rest of this work