Waste Eye

• 

  Recommended disposal instructions.

• 

  Webcam image submission.

Inspiration

Notes apps, surprisingly. As university students, notes apps have saved us a lot of paper, and this got us thinking about better waste management.

What it does

Waste Eye uses Google Cloud Vision and GPT-3.5 to identify objects and recommend methods of disposal.

How we built it

Most of the codebase is in JavaScript and Node.js, with Express.js handling backend API calls to Vision and GPT.

Challenges we ran into

As this was the first time we had worked with Artificial Intelligence APIs, getting everything to communicate was a challenge. We also ran into some merge conflicts on Github...and had some API keys end up on the web.

Accomplishments that we're proud of

We're pretty surprised we got it to work, actually. Getting to see Vision correctly identify most household items in our testing was really exciting to see!

What we learned

We learned a lot about authentication and version control, from both Google Cloud API and Github.

What's next for Waste Eye

A mobile app and better AI for waste management are high priorities for improving Waste Eye's effectiveness. GPT is not optimized for waste management and tends to provide general answers, while a library we can train would be more concise and thorough—and likely more helpful.

Built With

• css3
• express.js
• google-cloud-vision
• html5
• node.js