This repo implements a minimalistic version of Neural Style Transfer (NST) using PyTorch

Sample Outputs

Below are some sample outputs generated by this implementation. The left image is the content, the middle is the style, and the right is the generated image.

Customization

You can adjust various parameters in the excecution to experiment with different aspects of style transfer, such as:

• Number of iterations
• Style weight
• Content weight
• Learning rate

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgements

• Leon A. Gatys, Alexander S. Ecker, and Matthias Bethge for their pioneering work in Neural Style Transfer.
• The PyTorch team for providing an excellent deep learning framework.