IRIS: Image Recognition and Identification for Slack

Inspiration

Over 350 000 Canadians in the workforce are living with blindness, reduced vision, or other vision disabilities. Advances in accessible hardware and software have made tasks like coding much easier for the visually impaired, however these individuals still face challenges integrating into the work team. Our vision is to create a workplace where persons with impaired vision are fully and effortlessly integrated into their teams, by making the collaborative tools they use fully accessible. This starts with the world's fastest growing communication platform, Slack. Slack messages are compatible with screen readers, however images in Slack lack alternative text so screen readers can get no information about them. This means that visual information in a Slack workspace is completely inaccessible to team members with blindness or reduced vision.

What it does

IRIS is a bot that can be added to a Slack workspace. When an image is uploaded, IRIS generates a descriptive caption and prints it to the chat in screen reader-accessible format.

How we built it

The backend is built in Javascript and Python, and we connected to the Slack API though Glitch. Caption generation is done by Computer Vision from Microsoft Azure.

Challenges we ran into

This process was filled with challenges. We started by learning Javascript syntax from scratch, learned how to use Glitch, and connected to Microsoft Azure for the first time. Integrating all of our pieces into a cohesive code project was also a challenge.

Accomplishments that we're proud of

We're most proud of creating an app that can actually be added to Slack. Creating a clean, fully-functional product from scratch was incredibly rewarding.