Edit model card

RecycleTree - Plastics Classification Model

Banner

RecycleTree is a project from CZ3002 Advanced Software Engineering in Nanyang Technological University. It aims to enable users to have a more informed recycling experience, from finding the nearest recycling bins, to checking whether the item they wish to recycle can indeed be recycled, to learning more about recycling and contamination in general.

The whole project can be found on GitHub

This image classification model in particular is to classify plastic trash items into the following classes:

  • CD
  • Drinking Straws
  • Plastic Bags
  • Plastic Clothes Hanger
  • Plastic Container/Bottle
  • Plastic Disposables
  • Plastic Packaging
  • Plastic Packaging with Foil
  • Styrofoam

Training Data

The training dataset had 9646 images across 9 classes, with each class having roughly the same distribution of images. The images were either scraped from Google image search, or obtained by ourselves in real life.

Training Procedure

As the purpose of this model was to act just as a proof of concept for quick prototyping of RecycleTree, I opted to use the fast.ai library and a simple model architecture of ResNet34. The training procedure is following the recommendations from fast.ai

Other Models

There are also other models in the RecycleTree model series:

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Spaces using pyesonekyaw/recycletree_plastic 6