Inspiration
On campus, we frequently stop in confusion at waste disposal stations. Does the wrap for my burrito go into the garbage? Or is it recycling, or the paper items bin? There had to be a way to harness technology to simplify waste disposal for everyone.
What it does
Fooder is a simple to use web app that allows you to upload a photo of an item, or take one on the spot, and leverages machine-learning image classification to detect and determine which disposal category the items in the image belong to. For certain wastes like batteries, Fooder searches for nearby locations which accept the waste to ensure proper disposal, such as fire stations, in an embedded Google Maps.
How we built it
We leveraged React and Chakra UI for the frontend and Flask for the backend. We utilized the Google Mediapipe library with the EfficientDet-Lite0 Object-detection model for image classification, HuggingFace for disposal categorization, and Google Maps Platform's Places Service for the embedded map.
Challenges we ran into
Integrating the Maps API and ensuring the UI runs smoothly because disposal categorization for multiple items takes a long time.
Accomplishments that we're proud of
Completing a working MVP of Fooder in 24 hours using a variety of frameworks and APIs none of us had full familiarity with.
What we learned
Reading API documentation is mind numbing but always worth the thrill at the end when it works!
What's next for Fooder
Refining the ML model for greater accuracy and specific categorization. Plus, a history feature that saves previous searches.
Log in or sign up for Devpost to join the conversation.