We wanted to build tools to assist design to make sure that it would be accessible to people with colour vision deficiencies as well. Additionally, we also decided to create an AI that filters images based on other examples of image filtering. Meet BabyYodaBrush.
What it does
It can accept images and simulate three forms of color deficiency (Deuteranopia, Tritanopia, and Protanopia), as well as alter images to make the colors more distinguishable for those color deficiencies (Daltonization)
How I built it
The interface is through a Telegram bot, built in python, and makes use of the Daltonize project for python (https://github.com/joergdietrich/daltonize)
Challenges I ran into
Image processing is tough......
Accomplishments that I'm proud of
We got it to work!!
What I learned
AI models take longer to train the more you stare at it.
What's next for BabyYodaBrush
Hopefully live filtering through a device camera, so that we don't have to keep taking/sending pictures!
try it out at @color_blind_bot on Telegram!