In late 2021 our team of graduates students in UC Berkeley’ MIDS program assembled to tackle a problem using machine learning, one that could benefit society.
After much market research, interviews with subject matter experts and guidance from our professors we came up with Descriptive World. A simple idea that we might be able to assist the over 2 billion people worldwide that suffer from vision impairments.
The idea was to build a system that could leverage advances in artificial intelligence, edge computing and mobile devices to provide the visually impaired some independence in a way that was both hands off and intuitive.
We developed several prototypes and architectures which would eventually evolve into the Descriptive World system.
Our first project was to see how we could assist the visually impaired with sorting clothes and selecting matching outfits. A task that is impossible to accomplish without at least the initial assistance of a visually capable person. From this was research was born “Project Outfits”.