Chainer implementation of "Perceptual Losses for Real-Time Style Transfer and Super-Resolution".
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.

Chainer implementation of "Perceptual Losses for Real-Time Style Transfer and Super-Resolution"

Fast artistic style transfer by using feed forward network.

checkout resize-conv branch which provides better result.

  • input image size: 1024x768
  • process time(CPU): 17.78sec (Core i7-5930K)
  • process time(GPU): 0.994sec (GPU TitanX)


$ pip install chainer


Download VGG16 model and convert it into smaller file so that we use only the convolutional layers which are 10% of the entire model.



Need to train one image transformation network model per one style target. According to the paper, the models are trained on the Microsoft COCO dataset.

python -s <style_image_path> -d <training_dataset_path> -g <use_gpu ? gpu_id : -1>


python <input_image_path> -m <model_path> -o <output_image_path> -g <use_gpu ? gpu_id : -1>

This repo has pretrained models as an example.

  • example:
python sample_images/tubingen.jpg -m models/composition.model -o sample_images/output.jpg


python sample_images/tubingen.jpg -m models/seurat.model -o sample_images/output.jpg

Transfer only style but not color (--keep_colors option)

python <input_image_path> -m <model_path> -o <output_image_path> -g <use_gpu ? gpu_id : -1> --keep_colors

A collection of pre-trained models

Fashizzle Dizzle created pre-trained models collection repository, chainer-fast-neuralstyle-models. You can find a variety of models.

Difference from paper

  • Convolution kernel size 4 instead of 3.
  • Training with batchsize(n>=2) causes unstable result.

No Backward Compatibility

Jul. 19, 2016

This version is not compatible with the previous versions. You can't use models trained by the previous implementation. Sorry for the inconvenience!




Codes written in this repository based on following nice works, thanks to the author.

  • chainer-gogh Chainer implementation of neural-style. I heavily referenced it.
  • chainer-cifar10 Residual block implementation is referred.