Why we built it/mission
Our mission is to collect data for the people about recycling, with this data, the people can compare how much is actually recycled over the overall bought stuff.
Inspiration
We realized that a lot valuable data is lost within the process of recycling. Moreover, a lot of people don't tend to recycle.
What it does
It detects when a plastic bottle goes past a camera on a conveyor belt, and it uses user keys to detect when a user throws something in the trash.
How we built it
Tensorflow models, flutter, firebase. A lot of front end components are self-coded.
Challenges we ran into
The package object_detection for tensorflow actually uses tensorflow v1 rather than v2, this made it really hard for the backend developer to code the tensorflow model. However, we used google colab to preform version control.
Accomplishments that we're proud of
Working together in a great environment.
What we learned
Version control is very important, and debugging takes 10x longer than expected.
What's next for Recycle It
More training data, and actual hardware.
What we wanted to accomplish and what we learned in this hackathon
We learned to use tools to create data images (.xml) files for the model. We wanted to accomplish real-time object detection.
Log in or sign up for Devpost to join the conversation.