Inspiration
Over 7 million people use screen readers in the US alone. One of our members, Jacky, works at a school and has seen first hand how web navigation can be difficult for those who have specific accessibility needs.
What it does
Our chrome extension scans pages for images that do not have alt text describing them, and then generates and inserts descriptions so screen-readers can describe the images.
How we built it
The application is split into a front/back end. The front-end gathers images that do not have alt text, and sends them by URL to our server. Our server forwards the image information to Microsoft's Computer Vision API and returns the generated alt text.
Challenges we ran into
- Choosing a method to generate captions. (There are many)
- Setting up language support. (Again, many pathways)
- Developing UI (html/css/js is hard)
Accomplishments that we're proud of
It is very satisfying to develop a project that is a little more 'niche' in its environment, and one that could be a great aid to those who use it.
What we learned
There are a lot of subtleties to be learned when working on a chrome extension like determining how to persist data and develop in that specific environment.
What's next for Alt-o-Matic
There are may images that are purely decorative that do not need alt-text. Adding alts to these is both a waste of resources and might confuse screen-readers. To fix this we could have an in-between step where we classify images as content-enough or not.
Log in or sign up for Devpost to join the conversation.