We learned about the existence of a garbage island the size of Texas floating in the ocean. After careful deliberation, we thought about how we would personally tackle this problem.
What it does
Our app and backend service leverage the most recent cameraX api and keras machine learning library to accurately identify different categories of waste.
How we built it
Used Android Studio and Kotlin to build a camera-based application for android devices 5.0+. Trained sophisticated deep learning models to classify waste using Tensorflow and keras.
Challenges we ran into
Lack of support and documentation for Android Studio cameraX.
Accomplishments that we're proud of
Accuracy of our model (95% on training data)
What we learned
It takes a lot of time to create rigorous models
What's next for Classy Garbage
Extend to automated waste sorting machines.