Garbage Classification using EfficientNet

This repository contains the implementation of a garbage classification system inspired by the research paper: Deep Learning-based Waste Detection in Natural and Urban Environments. The system achieves efficient and accurate classification of waste into predefined categories using a deep learning approach.

Overview

Garbage pollution is a significant environmental challenge, and efficient waste classification is crucial for effective recycling and waste management. This project implements a model based on EfficientNet architecture, which achieves 83% classification accuracy, making it a practical tool for real-world applications such as smart waste bins and urban waste monitoring.

Features

• Deep Learning Backbone: Utilizes EfficientNet for image classification.
• Dataset Integration: Prepares and trains on the Kaggle Garbage Classification dataset.
• Accuracy: Achieves high accuracy with robust performance in diverse environmental conditions.
• Customizable: Easy-to-use training and testing pipelines.

Getting Started

Prerequisites

• Python 3.8 or higher
• TensorFlow / PyTorch
• NumPy
• OpenCV (for image preprocessing)

Installation

1. Clone this repository:

   git clone https://github.com/yourusername/garbage-classification
   cd garbage-classification

2. Install the dependencies:

   pip install -r requirements.txt

3. Download the dataset:

   Download the Garbage Classification Dataset from Kaggle: https://www.kaggle.com/datasets/asdasdasasdas/garbage-classification/data. Extract the dataset and place it in the data/ directory.

4. Preprocess the dataset:

   python preprocess_data.py --dataset-path data/Garbage\ Classification

Training the Model

Train the EfficientNet model using the provided dataset:

python train.py --dataset-path data/processed --epochs 50 --batch-size 32

Testing the Model

Evaluate the trained model on the test set:

python test.py --model-path models/efficientnet_best.pth --dataset-path data/processed

Implementation Details

Model Architecture

The project uses EfficientNet-B2, chosen for its balance of computational efficiency and accuracy. The architecture includes:

1. Pre-trained EfficientNet backbone for feature extraction.
2. Fully connected layers for waste category classification.

Waste Categories

The Kaggle dataset includes the following waste categories:

• Cardboard
• Glass
• Metal
• Paper
• Plastic
• Trash

These categories are used directly in the model.

Results

• Classification Accuracy: 83% on the test set.
• Dataset Size: The Kaggle dataset contains over 2,500 labeled images of waste.

Contributing

We welcome contributions! Please follow the steps below to contribute:

1. Fork the repository.
2. Create a feature branch (git checkout -b feature-name).
3. Commit your changes (git commit -m 'Add feature').
4. Push to the branch (git push origin feature-name).
5. Open a pull request.

Acknowledgments

• Dataset: This project uses the Garbage Classification dataset from Kaggle.
• Research Paper: Inspired by Deep Learning-based Waste Detection in Natural and Urban Environments.