Data Augmentation with DCGAN on ELPV Dataset

Overview

This repository contains a Jupyter notebook that demonstrates data augmentation using a Deep Convolutional Generative Adversarial Network (DCGAN). The notebook is designed to augment the ELPV dataset, specifically focusing on generating high-quality images of solar cells.

Objectives

1. Load images from the ELPV dataset.
2. Define and implement the DCGAN architecture.
3. Train the DCGAN model:
   • Track and save generator and discriminator loss after each epoch.
   • Save model checkpoints every 10 epochs.
   • Generate and save images after every 10 epochs.
4. Use the trained generator to create new images.

Setup

Prerequisites

• Python 3.x
• TensorFlow 2.x
• Matplotlib
• Numpy
• PIL
• ImageIO

Installation

Clone this repository and install the required packages:

git clone <repository-url>
cd <repository-directory>
pip install -r requirements.txt

Dataset

Ensure you have the ELPV dataset organized in a directory. The notebook expects the following directory structure:

/kaggle/input/elpv-classwise/1/

Usage

Open the notebook DCGAN-final-1.ipynb in Jupyter or any compatible environment and run the cells sequentially. Adjust the path to your dataset accordingly.

Results

After training, the model will generate new images simulating the input dataset, which will be saved in the specified output directory.

License

Specify the license under which the code is made available.

Contributors

List the contributors to the project.