Contrastive Learning for Unpaired Image-to-Image Translation

TL;DR

This paper introduces a contrastive learning framework for unpaired image-to-image translation that maximizes mutual information between corresponding patches, improving translation quality and efficiency.

Contribution

It presents a novel patch-based contrastive learning method that operates within individual images and enables effective unpaired translation with minimal data.

Findings

Enhances image translation quality

Reduces training time

Works with single-image domains

Abstract

In image-to-image translation, each patch in the output should reflect the content of the corresponding patch in the input, independent of domain. We propose a straightforward method for doing so -- maximizing mutual information between the two, using a framework based on contrastive learning. The method encourages two elements (corresponding patches) to map to a similar point in a learned feature space, relative to other elements (other patches) in the dataset, referred to as negatives. We explore several critical design choices for making contrastive learning effective in the image synthesis setting. Notably, we use a multilayer, patch-based approach, rather than operate on entire images. Furthermore, we draw negatives from within the input image itself, rather than from the rest of the dataset. We demonstrate that our framework enables one-sided translation in the unpaired image-to-image translation setting, while improving quality and reducing training time. In addition, our method can even be extended to the training setting where each "domain" is only a single image.