Neural Style Transfer Project

Overview

This project implements neural style transfer using TensorFlow and a pre-trained VGG19 model. The goal is to combine the content of one image with the artistic style of another, creating visually appealing results.

Demo

Sl no.	Content	Style	Output
1
2

Installation

1. Clone the repository:

   git clone https://github.com/yash-shimpi/DashToon-Generative-AI-Engineer

2. Install the required dependencies:

   pip install -r requirements.txt

3. Download the pre-trained VGG19 weights exluding top layer file from link and place it in the project root directory. (optional, else used Keras API)

Usage

1. Modify the content_path and style_path variables in the script to specify the paths of your content and style images.

2. Run the script:

   python3 notebook_script.py

3. View the output images and the training progress in the console.

Project Structure

• notebook_script.py: Main script for style transfer.
• notebook.ipynb: Main Jupyter notebook for style transfer.
• README.md: Project documentation.
• requirements.txt: List of project dependencies.
• vgg19_weights_tf_dim_ordering_tf_kernels_notop.h5: Pre-trained VGG19 weights file (downloaded separately).

Dependencies

• TensorFlow
• NumPy
• Matplotlib
• OpenCV
• PIL (Python Imaging Library)

Install dependencies using:

pip install -r requirements.txt