Unsupervised Cross-Domain Image Generation

Abstract

We study the problem of transferring a sample in one domain to an analog sample in another domain. Given two related domains, S and T, we would like to learn a generative function G that maps an input sample from S to the domain T, such that the output of a given function f, which accepts inputs in either domains, would remain unchanged. Other than the function f, the training data is unsupervised and consist of a set of samples from each domain. The Domain Transfer Network (DTN) we present employs a compound loss function that includes a multiclass GAN loss, an f-constancy component, and a regularizing component that encourages G to map samples from T to themselves. We apply our method to visual domains including digits and face images and demonstrate its ability to generate convincing novel images of previously unseen entities, while preserving their identity.

References in corpus (6)

• Conditional Generative Adversarial Nets
• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• Texture Networks: Feed-forward Synthesis of Textures and Stylized Images
• Neural Photo Editing with Introspective Adversarial Networks
• Image Super-Resolution Using Deep Convolutional Networks
• Inverting face embeddings with convolutional neural networks

Cited by in corpus (156)

• Deep Visual Domain Adaptation: A Survey
• Learning to Discover Cross-Domain Relations with Generative Adversarial Networks
• Exposing DeepFake Videos By Detecting Face Warping Artifacts
• Style Transfer from Non-Parallel Text by Cross-Alignment
• CyCADA: Cycle-Consistent Adversarial Domain Adaptation
• Unsupervised Image-to-Image Translation Networks
• Multimodal Unsupervised Image-to-Image Translation
• Cross-view Semantic Segmentation for Sensing Surroundings
• Fader Networks: Manipulating Images by Sliding Attributes
• Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World
• Unsupervised Machine Translation Using Monolingual Corpora Only
• DRIT++: Diverse Image-to-Image Translation via Disentangled Representations
• In Ictu Oculi: Exposing AI Generated Fake Face Videos by Detecting Eye Blinking
• Parallel-Data-Free Voice Conversion Using Cycle-Consistent Adversarial Networks
• Unsupervised Domain Adaptation through Self-Supervision
• An Introduction to Image Synthesis with Generative Adversarial Nets
• Toward Controlled Generation of Text
• Video-to-Video Synthesis
• Unsupervised Domain Adaptation with Adversarial Residual Transform Networks
• Generating Diverse High-Fidelity Images with VQ-VAE-2
• One-Shot Unsupervised Cross Domain Translation
• InstaGAN: Instance-aware Image-to-Image Translation
• Adversarial Multi-task Learning for Text Classification
• Fashion-Gen: The Generative Fashion Dataset and Challenge
• Triangle Generative Adversarial Networks
• Domain Stylization: A Strong, Simple Baseline for Synthetic to Real Image Domain Adaptation
• DNA-GAN: Learning Disentangled Representations from Multi-Attribute Images
• Pedestrian-Synthesis-GAN: Generating Pedestrian Data in Real Scene and Beyond
• One-Sided Unsupervised Domain Mapping
• Label Efficient Learning of Transferable Representations across Domains and Tasks
• Detecting GAN generated Fake Images using Co-occurrence Matrices
• Multiple-Attribute Text Style Transfer
• A Closed-form Solution to Photorealistic Image Stylization
• Adversarial Dropout Regularization
• Three-dimensional Generative Adversarial Nets for Unsupervised Metal Artifact Reduction
• Virtual Mixup Training for Unsupervised Domain Adaptation
• No-reference Screen Content Image Quality Assessment with Unsupervised Domain Adaptation
• Augmented Cyclic Adversarial Learning for Low Resource Domain Adaptation
• CalliGAN: Style and Structure-aware Chinese Calligraphy Character Generator
• Minimal-Entropy Correlation Alignment for Unsupervised Deep Domain Adaptation
• FaceLift: A transparent deep learning framework to beautify urban scenes
• BEV-Seg: Bird's Eye View Semantic Segmentation Using Geometry and Semantic Point Cloud
• Source Free Domain Adaptation with Image Translation
• Landmark Assisted CycleGAN for Cartoon Face Generation
• ELEGANT: Exchanging Latent Encodings with GAN for Transferring Multiple Face Attributes
• SPIGAN: Privileged Adversarial Learning from Simulation
• 3D-A-Nets: 3D Deep Dense Descriptor for Volumetric Shapes with Adversarial Networks
• DCAN: Dual Channel-wise Alignment Networks for Unsupervised Scene Adaptation
• Open Set Domain Adaptation by Backpropagation
• Gang of GANs: Generative Adversarial Networks with Maximum Margin Ranking
• Conditional Image-to-Image Translation
• WaveCycleGAN2: Time-domain Neural Post-filter for Speech Waveform Generation
• Separating Style and Content for Generalized Style Transfer
• Geometry-Consistent Generative Adversarial Networks for One-Sided Unsupervised Domain Mapping
• VIGAN: Missing View Imputation with Generative Adversarial Networks
• Unsupervised Domain-Specific Deblurring via Disentangled Representations
• Learning Face Age Progression: A Pyramid Architecture of GANs
• Twin-GAN -- Unpaired Cross-Domain Image Translation with Weight-Sharing GANs
• Developing Personalized Models of Blood Pressure Estimation from Wearable Sensors Data Using Minimally-trained Domain Adversarial Neural Networks
• RL-CycleGAN: Reinforcement Learning Aware Simulation-To-Real
• Domain-Invariant Adversarial Learning for Unsupervised Domain Adaption
• Facelet-Bank for Fast Portrait Manipulation
• GlyphGAN: Style-Consistent Font Generation Based on Generative Adversarial Networks
• Thermal Infrared Colorization via Conditional Generative Adversarial Network
• Transfer Learning for Context-Aware Question Matching in Information-seeking Conversations in E-commerce
• SharinGAN: Combining Synthetic and Real Data for Unsupervised Geometry Estimation
• Aspect-augmented Adversarial Networks for Domain Adaptation
• Adversarial Alignment of Class Prediction Uncertainties for Domain Adaptation
• Detection, Attribution and Localization of GAN Generated Images
• Geometry-Aware Symmetric Domain Adaptation for Monocular Depth Estimation
• Image-to-image Translation via Hierarchical Style Disentanglement
• DART: Domain-Adversarial Residual-Transfer Networks for Unsupervised Cross-Domain Image Classification
• DG-Font: Deformable Generative Networks for Unsupervised Font Generation
• Known-class Aware Self-ensemble for Open Set Domain Adaptation
• Learning Spatial Pyramid Attentive Pooling in Image Synthesis and Image-to-Image Translation
• Mix and match networks: encoder-decoder alignment for zero-pair image translation
• Few-shot Video-to-Video Synthesis
• Sim-Real Joint Reinforcement Transfer for 3D Indoor Navigation
• Unsupervised Multi-Domain Image Translation with Domain-Specific Encoders/Decoders
• Reciprocal Translation between SAR and Optical Remote Sensing Images with Cascaded-Residual Adversarial Networks
• Open-Edit: Open-Domain Image Manipulation with Open-Vocabulary Instructions
• GANwriting: Content-Conditioned Generation of Styled Handwritten Word Images
• CDCGen: Cross-Domain Conditional Generation via Normalizing Flows and Adversarial Training
• Exploring Explicit Domain Supervision for Latent Space Disentanglement in Unpaired Image-to-Image Translation
• Unsupervised Domain Adaptation for Visual Navigation
• A Unified Framework for Generalizable Style Transfer: Style and Content Separation
• TSIT: A Simple and Versatile Framework for Image-to-Image Translation
• ManiGAN: Text-Guided Image Manipulation
• Multi-Target Domain Adaptation via Unsupervised Domain Classification for Weather Invariant Object Detection
• Judge the Judges: A Large-Scale Evaluation Study of Neural Language Models for Online Review Generation
• A Theory of Output-Side Unsupervised Domain Adaptation
• Reinforcement Learning with Videos: Combining Offline Observations with Interaction
• Unsupervised Shadow Removal Using Target Consistency Generative Adversarial Network
• DotSCN: Group Re-identification via Domain-Transferred Single and Couple Representation Learning
• WarpGAN: Automatic Caricature Generation
• Detecting cognitive impairments by agreeing on interpretations of linguistic features
• Factorized Adversarial Networks for Unsupervised Domain Adaptation
• Polite Dialogue Generation Without Parallel Data
• Attention-GAN for Object Transfiguration in Wild Images
• Unpaired Photo-to-Caricature Translation on Faces in the Wild
• XOGAN: One-to-Many Unsupervised Image-to-Image Translation
• Cycle-consistent Conditional Adversarial Transfer Networks
• Generative Adversarial Network with Multi-Branch Discriminator for Cross-Species Image-to-Image Translation
• Learning Continuous Face Age Progression: A Pyramid of GANs
• r-BTN: Cross-domain Face Composite and Synthesis from Limited Facial Patches
• Cross Domain Knowledge Learning with Dual-branch Adversarial Network for Vehicle Re-identification
• Disentangled Cycle Consistency for Highly-realistic Virtual Try-On
• Point Cloud Completion by Learning Shape Priors
• CADA: Multi-scale Collaborative Adversarial Domain Adaptation for Unsupervised Optic Disc and Cup Segmentation
• Learning a Domain-Agnostic Visual Representation for Autonomous Driving via Contrastive Loss
• Image-to-Image Translation with Multi-Path Consistency Regularization
• Cross-Domain Adversarial Auto-Encoder
• From Shadow Generation to Shadow Removal
• Unsupervised Vehicle Re-identification with Progressive Adaptation
• Learning Domain Adaptive Features with Unlabeled Domain Bridges
• Joint Learning of Generative Translator and Classifier for Visually Similar Classes
• MangaGAN: Unpaired Photo-to-Manga Translation Based on The Methodology of Manga Drawing
• Attribute-Driven Spontaneous Motion in Unpaired Image Translation
• Cross Domain Image Generation through Latent Space Exploration with Adversarial Loss
• Domain-Adversarial Training of Self-Attention Based Networks for Land Cover Classification using Multi-temporal Sentinel-2 Satellite Imagery
• Air-Writing Translater: A Novel Unsupervised Domain Adaptation Method for Inertia-Trajectory Translation of In-air Handwriting
• Unsupervised Typography Transfer
• Memory-guided Unsupervised Image-to-image Translation
• Auto-Encoding for Shared Cross Domain Feature Representation and Image-to-Image Translation
• Unsupervised Representation Adversarial Learning Network: from Reconstruction to Generation
• Cross Domain Knowledge Transfer for Unsupervised Vehicle Re-identification
• Semantically Robust Unpaired Image Translation for Data with Unmatched Semantics Statistics
• Learning Classifiers for Domain Adaptation, Zero and Few-Shot Recognition Based on Learning Latent Semantic Parts
• GANILLA: Generative Adversarial Networks for Image to Illustration Translation
• Guiding the One-to-one Mapping in CycleGAN via Optimal Transport
• Generative Creativity: Adversarial Learning for Bionic Design
• Sequential Gating Ensemble Network for Noise Robust Multi-Scale Face Restoration
• Multi-Channel Auto-Encoders and a Novel Dataset for Learning Domain Invariant Representations of Histopathology Images
• A Survey on Adversarial Image Synthesis
• Learn to synthesize and synthesize to learn
• Profile to Frontal Face Recognition in the Wild Using Coupled Conditional GAN
• Improving Shape Deformation in Unsupervised Image-to-Image Translation
• Enlisting 3D Crop Models and GANs for More Data Efficient and Generalizable Fruit Detection
• Unsupervised Image-to-Image Translation with Stacked Cycle-Consistent Adversarial Networks
• Localising In Complex Scenes Using Balanced Adversarial Adaptation
• Sym-parameterized Dynamic Inference for Mixed-Domain Image Translation
• Online Domain Adaptation for Continuous Cross-Subject Liver Viability Evaluation Based on Irregular Thermal Data
• MakeupBag: Disentangling Makeup Extraction and Application
• Learning to Transfer Visual Effects from Videos to Images
• Video-to-Video Translation for Visual Speech Synthesis
• Image Manipulation with Perceptual Discriminators
• PSGAN++: Robust Detail-Preserving Makeup Transfer and Removal
• Head2HeadFS: Video-based Head Reenactment with Few-shot Learning
• MCMI: Multi-Cycle Image Translation with Mutual Information Constraints
• Towards Purely Unsupervised Disentanglement of Appearance and Shape for Person Images Generation
• Applications of Generative Adversarial Models in Visual Search Reformulation
• Spectra2pix: Generating Nanostructure Images from Spectra
• DGST : Discriminator Guided Scene Text detector
• Towards Visible and Thermal Drone Monitoring with Convolutional Neural Networks
• Generative Adversarial Networks with Inverse Transformation Unit
• Domain Adaptation with Incomplete Target Domains