Arbitrary Style Transfer using Graph Instance Normalization

TL;DR

This paper introduces Graph Instance Normalization (GrIN), a learnable normalization method using graph convolutional networks to improve style transfer by considering relationships between instances, enhancing robustness and applicability.

Contribution

The paper proposes a novel Graph Instance Normalization technique that captures relationships between instances for improved style transfer and other image translation tasks.

Findings

GrIN improves style transfer robustness.

Applicable to image translation and domain adaptation.

Enhances global style learning.

Abstract

Style transfer is the image synthesis task, which applies a style of one image to another while preserving the content. In statistical methods, the adaptive instance normalization (AdaIN) whitens the source images and applies the style of target images through normalizing the mean and variance of features. However, computing feature statistics for each instance would neglect the inherent relationship between features, so it is hard to learn global styles while fitting to the individual training dataset. In this paper, we present a novel learnable normalization technique for style transfer using graph convolutional networks, termed Graph Instance Normalization (GrIN). This algorithm makes the style transfer approach more robust by taking into account similar information shared between instances. Besides, this simple module is also applicable to other tasks like image-to-image translation or domain adaptation.