Ever heard of the saying "one bad apple spoils the whole barrel"! Well with Bad Apples we will detect the Bad Apples before they spoil the whole barrel. Bad Apples runs on a YoloV8 model that is trained on 1000+ images that are defined as either "Fresh" or "Rotten" apples. Due to our dataset being relatively small compared to others, it only trained in just over an hour. Bad Apples has potential use in supermarkets where we can automatically alert storekeepers as to whether or not their product is safe to sell. We struggled to use react and flutter for the first 6 hours of work, and nearly gave up several times.
Inspired
We have personal struggles with groceries spoiling and wanted to create something that could reduce food waste
How we built it
The backend was created in python and used flask to take input from the user. Trained pytorch model in google collab with ultralytics.
Challenges
Attempting to create a program using a framework(React) that we were unfamiliar with. Our greatest challenge was integrating the model to accept web cam input and display back to user.
one apple was harmed in the making of this program
Log in or sign up for Devpost to join the conversation.