Waste Classifier

A convolutional neural network that classifies waste based on whether it is recyclable or organic.

Table of Contents

• About
• Features
• Screenshots
• Installation
• Usage
• Model Training
• Contributing
• License

About

The Waste Classifier is a deep learning model built for the purpose of waste classification. It utilizes a custom convolutional neural network (CNN) to classify waste into two categories: organic and recyclable. This is a personal deep learning project using PyTorch that aims to contribute to waste management initiatives by automating the sorting process, impacting the environment for the better! :)

Features

• Convolutional Neural Network for waste classification.
• Binary classification: Organic and Recyclable.
• Model training script for customization.
• Model testing script for predictions with a trained model

Screenshots

An example of prediction results from a trained model with an average of 90% accuracy:

Installation

To get started, follow these steps:

1. Clone the repository:

   git clone https://github.com/alvaroderas/waste-classifier.git

2. Install dependencies:

   pip install -r requirements.txt

Model Training

1. Edit any constants to experiment with training.

2. Navigate to src directory and run the model training script:

   python train.py

Usage

To use the Waste Classifier to visualize waste classification predictions with a trained model, navigate to test directory and run the script:

python predictions.py

License

This project is licensed under the MIT License.