image-colorization

Overview

This project demonstrates how to colorize grayscale images using a Convolutional Neural Network (CNN) trained on the CIFAR-10 dataset. The model learns to predict the a and b color channels from the L (grayscale) channel in the CIELAB color space.

Features

• Loads and preprocesses the CIFAR-10 dataset
• Converts RGB images to LAB color space
• Trains a CNN to predict color channels from grayscale inputs
• Visualizes colorization results using Matplotlib
• Saves the trained model for reuse

Environment

This project was developed and trained in Google Colab, utilizing an NVIDIA A100 GPU to speed up training. Make sure GPU acceleration is enabled in your Colab runtime for optimal performance.

Dependencies

pip install tensorflow opencv-python matplotlib scikit-image Pillow numpy

How to Run

1. Open the notebook in Google Colab
2. Make sure GPU runtime is enabled: Runtime -> Change runtime type -> Hardware accelerator -> GPU
3. Run all cells step-by-step