One-Sided Unsupervised Domain Mapping

Abstract

In unsupervised domain mapping, the learner is given two unmatched datasets $A$ and $B$. The goal is to learn a mapping $G_{AB}$ that translates a sample in $A$ to the analog sample in $B$. Recent approaches have shown that when learning simultaneously both $G_{AB}$ and the inverse mapping $G_{BA}$, convincing mappings are obtained. In this work, we present a method of learning $G_{AB}$ without learning $G_{BA}$. This is done by learning a mapping that maintains the distance between a pair of samples. Moreover, good mappings are obtained, even by maintaining the distance between different parts of the same sample before and after mapping. We present experimental results that the new method not only allows for one sided mapping learning, but also leads to preferable numerical results over the existing circularity-based constraint. Our entire code is made publicly available at https://github.com/sagiebenaim/DistanceGAN .

to be published in NIPS 2017

References in corpus (7)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
• FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation
• Texture Networks: Feed-forward Synthesis of Textures and Stylized Images
• Unsupervised Cross-Domain Image Generation
• DualGAN: Unsupervised Dual Learning for Image-to-Image Translation
• Least Squares Generative Adversarial Networks

Cited by in corpus (16)

• How Generative Adversarial Networks and Their Variants Work: An Overview
• Multimodal Unsupervised Image-to-Image Translation
• A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications
• An Introduction to Image Synthesis with Generative Adversarial Nets
• One-Shot Unsupervised Cross Domain Translation
• Image-Image Domain Adaptation with Preserved Self-Similarity and Domain-Dissimilarity for Person Re-identification
• Generative adversarial interpolative autoencoding: adversarial training on latent space interpolations encourage convex latent distributions
• Geometry-Consistent Generative Adversarial Networks for One-Sided Unsupervised Domain Mapping
• VIGAN: Missing View Imputation with Generative Adversarial Networks
• Zero-shot Domain Adaptation without Domain Semantic Descriptors
• Joint one-sided synthetic unpaired image translation and segmentation for colorectal cancer prevention
• Attention-GAN for Object Transfiguration in Wild Images
• Specular-to-Diffuse Translation for Multi-View Reconstruction
• Visual Analogies between Atari Games for Studying Transfer Learning in RL
• Image Manipulation with Perceptual Discriminators
• Synthesizing Photorealistic Images with Deep Generative Learning