Image-to-Image Translation with Conditional Adversarial Networks

TL;DR

This paper presents a versatile conditional adversarial network framework for image-to-image translation that learns both the mapping and the loss function, enabling diverse applications without task-specific loss engineering.

Contribution

The authors introduce a general-purpose conditional adversarial network approach that automates learning the loss function for image translation tasks, simplifying the process across various applications.

Findings

Effective at synthesizing photos from label maps

Able to reconstruct objects from edge maps

Capable of colorizing images

Abstract

We investigate conditional adversarial networks as a general-purpose solution to image-to-image translation problems. These networks not only learn the mapping from input image to output image, but also learn a loss function to train this mapping. This makes it possible to apply the same generic approach to problems that traditionally would require very different loss formulations. We demonstrate that this approach is effective at synthesizing photos from label maps, reconstructing objects from edge maps, and colorizing images, among other tasks. Indeed, since the release of the pix2pix software associated with this paper, a large number of internet users (many of them artists) have posted their own experiments with our system, further demonstrating its wide applicability and ease of adoption without the need for parameter tweaking. As a community, we no longer hand-engineer our mapping functions, and this work suggests we can achieve reasonable results without hand-engineering our loss functions either.