MXNet-Gluon-Style-Transfer

This repo provides MXNet Implementation of Neural Style Transfer and MSG-Net.

Tabe of content

• Slow Neural Style Transfer
• Real-time Style Transfer
  • Stylize Images using Pre-trained MSG-Net
  • Train Your Own MSG-Net Model

Neural Style

A Neural Algorithm of Artistic Style by Leon A. Gatys, Alexander S. Ecker, and Matthias Bethge.

Download the images

python download_images.py 

Neural style transfer

python main.py optim --content-image images/content/venice-boat.jpg --style-image images/styles/candy.jpg

• --content-image: path to content image.
• --style-image: path to style image.
• --output-image: path for saving the output image.
• --content-size: the content image size to test on.
• --style-size: the style image size to test on.
• --cuda: set it to 1 for running on GPU, 0 for CPU.

Real-time Style Transfer

Stylize Images Using Pre-trained MSG-Net

0. Download the images and pre-trained model

   python download_images.py 
   	python models/download_model.py
   

1. Test the model

   	python main.py eval --content-image images/content/venice-boat.jpg --style-image images/styles/candy.jpg --model models/21styles.params --content-size 1024
   

• If you don't have a GPU, simply set --cuda=0. For a different style, set --style-image path/to/style. If you would to stylize your own photo, change the --content-image path/to/your/photo. More options:

  • --content-image: path to content image you want to stylize.
  • --style-image: path to style image (typically covered during the training).
  • --model: path to the pre-trained model to be used for stylizing the image.
  • --output-image: path for saving the output image.
  • --content-size: the content image size to test on.
  • --cuda: set it to 1 for running on GPU, 0 for CPU.

Train Your Own MSG-Net Model

0. Download the style images and COCO dataset Note: Dataset from COCO 2014. The dataset annotations and site are Copyright COCO Consortium and licensed CC BY 4.0 Attribution. The images within the dataset are available under the Flickr Terms of Use. See original dataset source for details

   python download_images.py 
   	python dataset/download_dataset.py
   

1. Train the model

   	python main.py train --epochs 4
   

• If you would like to customize styles, set --style-folder path/to/your/styles. More options:
  • --style-folder: path to the folder style images.
  • --vgg-model-dir: path to folder where the vgg model will be downloaded.
  • --save-model-dir: path to folder where trained model will be saved.
  • --cuda: set it to 1 for running on GPU, 0 for CPU.

The code is mainly modified from PyTorch-Style-Transfer.