UNET-Image-Colorization

Overview

This project uses a deep learning model to colorize black and white images. It use a U-Net architecture and is particularly focused on colorizing images of dogs. The model is trained and tested using the Linnaeus 5 dataset.

Prerequisites

• Python 3.x
• Pip (Python package manager)

Installation

Clone this repository and install the required Python packages:

git clone https://github.com/NathanNT/GAN-Image-Colorization
cd GAN-Image-Colorization
pip install -r requirements.txt

Dataset

You need to download the Linnaeus 5 dataset in 64x64 and structure it as follows:

• Organize the images in test/dog and train/dog directories within the project folder.

You can find the Linnaeus 5 dataset at Linnaeus 5 Dataset or an alternative source. Ensure to adhere to the dataset's terms of use.

Running the Application

Execute the main script app.py to start the training and colorization process:

python app.py

The script processes the images, trains the model. Each epoch generates a colorized image from the test set, shown alongside its original and grayscale versions. The model's output images are saved in an output directory.