CycleGAN Monet-to-Photo Translation

Turn a Monet-style painting into a photo

Released in 2017, this model exploits a novel technique for image translation, in which two models translating from A to B and vice versa are trained jointly with adversarial training. In addition to the adversarial loss, cycle consistency is also enforced in the loss function: when the output of the first translator is fed into the second, the final result is encouraged to match the input of the first translator. This allows successful training for image translation tasks in which only unpaired training data can be collected. This model was trained to translate Monet-style paintings into photos.

Number of layers: 94 | Parameter count: 2,855,811 | Trained size: 12 MB |

Training Set Information

A collection of Monet paintings and photos downloaded from Flickr.

Examples

Download Example Notebook

Open in Wolfram Cloud

Resource retrieval

Get the pre-trained net:

In[1]:=

Adapt to any size

Automatic image resizing can be avoided by replacing the net encoders. First get the net:

In[3]:=

Out[5]=

Out[6]=

Out[8]=

$NetInformation[ NetModel["CycleGAN Monet-to-Photo Translation"], \ "ArraysTotalElementCount"]$

Out[9]=

Out[10]=

Out[11]=

Out[14]=

Out[16]=

Construction Notebook

Download Construction Notebook

Open in Wolfram Cloud

Requirements

Wolfram Language 11.3 (March 2018) or above

Resource History

Date Created: 12 February 2018
Latest Update: 21 June 2018

Reference

J. Zhu, T. Park, P. Isola, A. A. Efros, "Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks," arXiv:1703.10593 (2017)
Available from: https://github.com/junyanz/CycleGAN
Rights: BSD 3-Clause "Modified" License