Guiding the One-to-one Mapping in CycleGAN via Optimal Transport
arXiv:1811.06284
Abstract
CycleGAN is capable of learning a one-to-one mapping between two data distributions without paired examples, achieving the task of unsupervised data translation. However, there is no theoretical guarantee on the property of the learned one-to-one mapping in CycleGAN. In this paper, we experimentally find that, under some circumstances, the one-to-one mapping learned by CycleGAN is just a random one within the large feasible solution space. Based on this observation, we explore to add extra constraints such that the one-to-one mapping is controllable and satisfies more properties related to specific tasks. We propose to solve an optimal transport mapping restrained by a task-specific cost function that reflects the desired properties, and use the barycenters of optimal transport mapping to serve as references for CycleGAN. Our experiments indicate that the proposed algorithm is capable of learning a one-to-one mapping with the desired properties.
arXiv commentThe Thirty-Third AAAI Conference on Artificial Intelligence (AAAI 2019)
References in corpus (14)
- Adam: A Method for Stochastic Optimization
- Conditional Generative Adversarial Nets
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
- Spectral Normalization for Generative Adversarial Networks
- Conditional Image Synthesis With Auxiliary Classifier GANs
- Improved Training of Wasserstein GANs
- Dual Learning for Machine Translation
- Unsupervised Image-to-Image Translation Networks
- Unsupervised Cross-Domain Image Generation
- On the regularization of Wasserstein GANs
- Generative Visual Manipulation on the Natural Image Manifold
- Large-Scale Optimal Transport and Mapping Estimation
- Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks
- Activation Maximization Generative Adversarial Nets