GAIT: Gradient Adjusted Unsupervised Image-to-Image Translation

TL;DR

This paper introduces GAIT, a novel unsupervised image-to-image translation method that preserves uniform regions by adjusting gradients, addressing artifacts caused by distribution mismatches.

Contribution

The paper proposes a gradient adjustment loss for unsupervised IIT, improving the preservation of uniform regions and reducing artifacts compared to existing adversarial methods.

Findings

Improved preservation of uniform regions in translated images.

Quantitative and qualitative performance gains over baseline methods.

Effective on datasets with different background distributions.

Abstract

Image-to-image translation (IIT) has made much progress recently with the development of adversarial learning. In most of the recent work, an adversarial loss is utilized to match the distributions of the translated and target image sets. However, this may create artifacts if two domains have different marginal distributions, for example, in uniform areas. In this work, we propose an unsupervised IIT method that preserves the uniform regions after the translation. The gradient adjustment loss, which is the L2 norm between the Sobel response of the target image and the adjusted Sobel response of the source images, is utilized. The proposed method is validated on the jellyfish-to-Haeckel dataset, which is prepared to demonstrate the mentioned problem, which contains images with different background distributions. We demonstrate that our method obtained a performance gain compared to the baseline method qualitatively and quantitatively, showing the effectiveness of the proposed method.