Invertible Conditional GANs for image editing

Abstract

Generative Adversarial Networks (GANs) have recently demonstrated to successfully approximate complex data distributions. A relevant extension of this model is conditional GANs (cGANs), where the introduction of external information allows to determine specific representations of the generated images. In this work, we evaluate encoders to inverse the mapping of a cGAN, i.e., mapping a real image into a latent space and a conditional representation. This allows, for example, to reconstruct and modify real images of faces conditioning on arbitrary attributes. Additionally, we evaluate the design of cGANs. The combination of an encoder with a cGAN, which we call Invertible cGAN (IcGAN), enables to re-generate real images with deterministic complex modifications.

Accepted paper at NIPS 2016 Workshop on Adversarial Training

Cited by in corpus (47)

• Fader Networks: Manipulating Images by Sliding Attributes
• A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications
• Triangle Generative Adversarial Networks
• Generative OpenMax for Multi-Class Open Set Classification
• GeneGAN: Learning Object Transfiguration and Attribute Subspace from Unpaired Data
• STGAN: A Unified Selective Transfer Network for Arbitrary Image Attribute Editing
• PA-GAN: Progressive Attention Generative Adversarial Network for Facial Attribute Editing
• ACE: Adapting to Changing Environments for Semantic Segmentation
• What Will Your Child Look Like? DNA-Net: Age and Gender Aware Kin Face Synthesizer
• MulGAN: Facial Attribute Editing by Exemplar
• Detection, Attribution and Localization of GAN Generated Images
• Editing in Style: Uncovering the Local Semantics of GANs
• Multiple-Identity Image Attacks Against Face-based Identity Verification
• Unsupervised Multi-Domain Image Translation with Domain-Specific Encoders/Decoders
• Fully-Featured Attribute Transfer
• MagGAN: High-Resolution Face Attribute Editing with Mask-Guided Generative Adversarial Network
• Multimodal Generative Models for Compositional Representation Learning
• Biphasic Learning of GANs for High-Resolution Image-to-Image Translation
• Conditional Image Generation and Manipulation for User-Specified Content
• Coconditional Autoencoding Adversarial Networks for Chinese Font Feature Learning
• MHVAE: a Human-Inspired Deep Hierarchical Generative Model for Multimodal Representation Learning
• Texture Deformation Based Generative Adversarial Networks for Face Editing
• Garment Design with Generative Adversarial Networks
• Disentanglement for Discriminative Visual Recognition
• Dual Generator Generative Adversarial Networks for Multi-Domain Image-to-Image Translation
• DualDis: Dual-Branch Disentangling with Adversarial Learning
• Emotion Generation and Recognition: A StarGAN Approach
• GAN-Based Facial Attractiveness Enhancement
• PriorGAN: Real Data Prior for Generative Adversarial Nets
• IntersectGAN: Learning Domain Intersection for Generating Images with Multiple Attributes
• Typeface Completion with Generative Adversarial Networks
• Learning Controllable Disentangled Representations with Decorrelation Regularization
• MU-GAN: Facial Attribute Editing based on Multi-attention Mechanism
• How to make a pizza: Learning a compositional layer-based GAN model
• Reconstructing faces from voices
• Face Attribute Invertion
• Fashion-AttGAN: Attribute-Aware Fashion Editing with Multi-Objective GAN
• Expression Conditional GAN for Facial Expression-to-Expression Translation
• Multi-component Image Translation for Deep Domain Generalization
• Learn to synthesize and synthesize to learn
• Adversarial Code Learning for Image Generation
• Nested Scale Editing for Conditional Image Synthesis
• 3D Dense Geometry-Guided Facial Expression Synthesis by Adversarial Learning
• Learning more expressive joint distributions in multimodal variational methods
• Generative Adversarial Networks for Multimodal Representation Learning in Video Hyperlinking
• Confidence-aware Adversarial Learning for Self-supervised Semantic Matching
• GL-GAN: Adaptive Global and Local Bilevel Optimization model of Image Generation