preetam-das26 / colorize-image

Colorize Images

This project focuses on using PyTorch to colorize black-and-white landscape images. It employs a convolutional neural network trained with 800 grayscale landscape images to achieve image colorization.

Features

• Deep Learning: Utilizes a convolutional neural network for image colorization.
• PyTorch: Implements the model and training using PyTorch.
• Image Processing: Handles grayscale to color image conversion.

Initial:

Final:

Installation

1. Clone this repository.
2. Run this code.
3. Create a python virtual environment and install the required packages.

   cd src
   python3 -W ignore color_img.py ../test.jpg ../out.jpg
   
   

Project Structure

• color_img.py: Script for colorizing images.
• train.py: Script for training the neural network model.
• model.py: Defines the neural network architecture.
• utils.py: Utility functions for data preprocessing and other tasks.
• requirements.txt: List of dependencies required for the project.

Topics

• Machine Learning
• Computer Vision
• PyTorch
• Image Colorization
• Convolutional Neural Networks