CycleGAN with Better Cycles
This work addresses image-to-image translation for researchers and practitioners, but it is incremental as it builds on existing CycleGAN methods.
The authors tackled the problem of unrealistic images in CycleGAN due to pixel-level cycle consistency, proposing three simple modifications that achieved better results with fewer artifacts.
CycleGAN provides a framework to train image-to-image translation with unpaired datasets using cycle consistency loss [4]. While results are great in many applications, the pixel level cycle consistency can potentially be problematic and causes unrealistic images in certain cases. In this project, we propose three simple modifications to cycle consistency, and show that such an approach achieves better results with fewer artifacts.