Multi-marginal Wasserstein GAN

Abstract

Multiple marginal matching problem aims at learning mappings to match a source domain to multiple target domains and it has attracted great attention in many applications, such as multi-domain image translation. However, addressing this problem has two critical challenges: (i) Measuring the multi-marginal distance among different domains is very intractable; (ii) It is very difficult to exploit cross-domain correlations to match the target domain distributions. In this paper, we propose a novel Multi-marginal Wasserstein GAN (MWGAN) to minimize Wasserstein distance among domains. Specifically, with the help of multi-marginal optimal transport theory, we develop a new adversarial objective function with inner- and inter-domain constraints to exploit cross-domain correlations. Moreover, we theoretically analyze the generalization performance of MWGAN, and empirically evaluate it on the balanced and imbalanced translation tasks. Extensive experiments on toy and real-world datasets demonstrate the effectiveness of MWGAN.

This paper is accepted by NeurIPS 2019

Cited by in corpus (9)

• Collaborative Unsupervised Domain Adaptation for Medical Image Diagnosis
• Dense Regression Network for Video Grounding
• Location-aware Graph Convolutional Networks for Video Question Answering
• PriorGAN: Real Data Prior for Generative Adversarial Nets
• Intelligent Home 3D: Automatic 3D-House Design from Linguistic Descriptions Only
• Online Adaptive Asymmetric Active Learning with Limited Budgets
• Improving Generative Adversarial Networks with Local Coordinate Coding
• Robust W-GAN-Based Estimation Under Wasserstein Contamination
• Unaligned Image-to-Image Translation by Learning to Reweight