CycleGAN: style transfer without paired data via cycle consistency

Released in: Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks



Image-to-image translation is a class of vision and graphics problems where the goal is to learn the mapping between an input image and an output image using a training set of aligned image pairs. However, for many tasks, paired training data will not be available. The authors present an approach for learning to translate an image from a source domain X to a target domain Y in the absence of paired examples. The main goal is to learn a mapping G: X → Y such that the distribution of images from G(X) is indistinguishable from the distribution Y using an adversarial loss. Because this mapping is highly under-constrained, the authors couple it with an inverse mapping F: Y → X and introduce a cycle consistency loss to push F(G(X)) ≈ X (and vice versa). Qualitative results are presented on several tasks where paired training data does not exist, including collection style transfer, object transfiguration, season transfer, photo enhancement, etc. Quantitative comparisons against several prior methods demonstrate the superiority of this approach.


Year Released

Key Links & Stats

CycleGAN and pix2pix in PyTorch


Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks

@inproceedings{CycleGAN2017, title={Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks}, author={Zhu, Jun-Yan and Park, Taesung and Isola, Phillip and Efros, Alexei A}, booktitle={Computer Vision (ICCV), 2017 IEEE International Conference on}, year={2017} }

ML Tasks

  1. General
  2. Domain Adaptation
  3. Style Transfer

ML Platform

  1. Pytorch


  1. General
  2. Still Image


  1. General

CG Platform

  1. Not Applicable

Related organizations

UC Berkeley