PyTorch Neural Image Style Transfer

Neural image style transfer is the technique of taking the style from one image and the content from another and merging them together into a cohesive image. This is done by using convolutional neural networks to extract representations of style and content from the respective images. These repesentations are then used to generate a new image in which the difference between the new image and these representations is minimized. The code developed here is based on the work done in:

L. A. Gatys, A. S. Ecker and M. Bethge, "Image Style Transfer Using Convolutional Neural Networks," 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, 2016, pp. 2414-2423.

The code also takes some inspiration from Neural Style Tansfer Using Pytorch by Winston Herring.

Setup and Run

Currently, the easiest way to get up and running with the image styler is to utilize the Docker image, but you can also clone the repository if you desire. A PyPi package will be setup soon!

Docker

1. Ensure Docker (or NVIDIA-Docker if you want to use a GPU) is installed
2. Run docker run -it -v $(pwd):/workspace james-mchugh/image-styler to get usage information for the container
3. As an example, if you want to use the content from content.jpg, the style from style.jpg, and output the image the output.jpg run the following command:

docker run -it -v $(pwd):/workspace james-mchugh/image-styler content.jpg
 style.jpg output.jpg

Source

1. Clone the repo
2. In the repository directory, install the Python package using pip install .
3. Run python -m image_styler to get usage information about the package
4. As an example, if you want to use the content from content.jpg, the style from style.jpg, and output the image the output.jpg run the following command:

python -m image_styler content.jpg style.jpg output.jpg

More examples and details to come!