Geometry-Consistent Generative Adversarial Networks for One-Sided Unsupervised Domain Mapping
arXiv:1809.05852
Abstract
Unsupervised domain mapping aims to learn a function to translate domain X to Y by a function GXY in the absence of paired examples. Finding the optimal GXY without paired data is an ill-posed problem, so appropriate constraints are required to obtain reasonable solutions. One of the most prominent constraints is cycle consistency, which enforces the translated image by GXY to be translated back to the input image by an inverse mapping GYX. While cycle consistency requires the simultaneous training of GXY and GY X, recent studies have shown that one-sided domain mapping can be achieved by preserving pairwise distances between images. Although cycle consistency and distance preservation successfully constrain the solution space, they overlook the special properties that simple geometric transformations do not change the semantic structure of images. Based on this special property, we develop a geometry-consistent generative adversarial network (GcGAN), which enables one-sided unsupervised domain mapping. GcGAN takes the original image and its counterpart image transformed by a predefined geometric transformation as inputs and generates two images in the new domain coupled with the corresponding geometry-consistency constraint. The geometry-consistency constraint reduces the space of possible solutions while keep the correct solutions in the search space. Quantitative and qualitative comparisons with the baseline (GAN alone) and the state-of-the-art methods including CycleGAN and DistanceGAN demonstrate the effectiveness of our method.
References in corpus (24)
- Adam: A Method for Stochastic Optimization
- Instance Normalization: The Missing Ingredient for Fast Stylization
- Generative Adversarial Text to Image Synthesis
- Conditional Image Generation with PixelCNN Decoders
- Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks
- Improved Techniques for Training GANs
- Toward Multimodal Image-to-Image Translation
- The Cityscapes Dataset for Semantic Urban Scene Understanding
- Adversarial Feature Learning
- Adversarially Learned Inference
- Texture Networks: Feed-forward Synthesis of Textures and Stylized Images
- Domain Separation Networks
- Unsupervised Image-to-Image Translation Networks
- Unsupervised Cross-Domain Image Generation
- Augmented CycleGAN: Learning Many-to-Many Mappings from Unpaired Data
- DualGAN: Unsupervised Dual Learning for Image-to-Image Translation
- Unsupervised Attention-guided Image to Image Translation
- One-Shot Unsupervised Cross Domain Translation
- One-Sided Unsupervised Domain Mapping
- Conditional Image-to-Image Translation
- Stereoscopic Neural Style Transfer
- Avatar-Net: Multi-scale Zero-shot Style Transfer by Feature Decoration
- Discriminative Region Proposal Adversarial Networks for High-Quality Image-to-Image Translation
- Mix and match networks: encoder-decoder alignment for zero-pair image translation