Invertible Conditional GANs for image editing

TL;DR

This paper introduces Invertible cGANs (IcGANs), enabling real image reconstruction and attribute-based editing by combining encoders with conditional GANs for deterministic modifications.

Contribution

It proposes a novel framework that integrates encoders with cGANs to invert the mapping, allowing real image editing and reconstruction.

Findings

Enables reconstruction of real images from latent space

Allows attribute-based image editing

Demonstrates deterministic modifications of real images

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.