Spades - Neural Style Transfer

Spades is a Neural Style Transfer (NST) project that applies artistic styles to images using deep learning. The project leverages a pre-trained VGG19 model to extract content and style features, blending them to create visually stunning artwork.

Features

• Uses VGG19 for feature extraction
• Transfers artistic styles from one image to another
• Supports high-resolution image processing
• Can be extended into a web-based application using Flask

Tech Stack

• Python (Primary language)
• TensorFlow / PyTorch (Deep learning framework)
• OpenCV (Image processing)
• Matplotlib (Visualization)
• Flask (Optional - for a web UI)

Installation

Ensure you have Python installed, then install dependencies:

pip install tensorflow torch torchvision opencv-python numpy matplotlib flask

Project Structure

Spades/
│── main.py                # Runs the NST process
│── style_transfer.py       # Core NST logic
│── utils.py                # Helper functions
│── models/                 # Pre-trained VGG19
│── images/                 # Input and output images
│── static/                 # (For Flask) Stores processed images
│── templates/              # (For Flask) HTML files for UI

Usage

Run the script to apply style transfer:

python main.py

Next Steps

• Implement optimization for style blending
• Add a Flask web interface
• Allow real-time NST with video processing

License

This project is open-source under the MIT License.