Multi-mapping Image-to-Image Translation via Learning Disentanglement

Abstract

Recent advances of image-to-image translation focus on learning the one-to-many mapping from two aspects: multi-modal translation and multi-domain translation. However, the existing methods only consider one of the two perspectives, which makes them unable to solve each other's problem. To address this issue, we propose a novel unified model, which bridges these two objectives. First, we disentangle the input images into the latent representations by an encoder-decoder architecture with a conditional adversarial training in the feature space. Then, we encourage the generator to learn multi-mappings by a random cross-domain translation. As a result, we can manipulate different parts of the latent representations to perform multi-modal and multi-domain translations simultaneously. Experiments demonstrate that our method outperforms state-of-the-art methods.

arXiv commentAccepted by NeurIPS 2019. Code will be available at https://github.com/Xiaoming-Yu/DMIT

Cited by in corpus (19)

• Swapping Autoencoder for Deep Image Manipulation
• StarGAN v2: Diverse Image Synthesis for Multiple Domains
• Sketch-Guided Scenery Image Outpainting
• Deep Image Spatial Transformation for Person Image Generation
• DeepFaceEditing: Deep Face Generation and Editing with Disentangled Geometry and Appearance Control
• Arbitrary Style Transfer via Multi-Adaptation Network
• Image-to-image Translation via Hierarchical Style Disentanglement
• README: REpresentation learning by fairness-Aware Disentangling MEthod
• Open-Edit: Open-Domain Image Manipulation with Open-Vocabulary Instructions
• TSIT: A Simple and Versatile Framework for Image-to-Image Translation
• Dual Attention GANs for Semantic Image Synthesis
• Multimodal Image-to-Image Translation via Mutual Information Estimation and Maximization
• Combining Attention with Flow for Person Image Synthesis
• Image Synthesis via Semantic Composition
• Semi-supervised Meta-learning with Disentanglement for Domain-generalised Medical Image Segmentation
• Toward Zero-Shot Unsupervised Image-to-Image Translation
• SpaceEdit: Learning a Unified Editing Space for Open-Domain Image Editing
• Unaligned Image-to-Image Translation by Learning to Reweight
• TransferI2I: Transfer Learning for Image-to-Image Translation from Small Datasets