Universal Style Transfer via Feature Transforms

TL;DR

This paper introduces a universal style transfer method using feature whitening and coloring transforms embedded in an image reconstruction network, enabling high-quality style transfer without pre-training on specific styles.

Contribution

The proposed method uniquely employs feature covariance matching with whitening and coloring transforms, eliminating the need for style-specific training.

Findings

Produces high-quality stylized images

Effectively generalizes to unseen styles

Visualizes feature transformations and textures

Abstract

Universal style transfer aims to transfer arbitrary visual styles to content images. Existing feed-forward based methods, while enjoying the inference efficiency, are mainly limited by inability of generalizing to unseen styles or compromised visual quality. In this paper, we present a simple yet effective method that tackles these limitations without training on any pre-defined styles. The key ingredient of our method is a pair of feature transforms, whitening and coloring, that are embedded to an image reconstruction network. The whitening and coloring transforms reflect a direct matching of feature covariance of the content image to a given style image, which shares similar spirits with the optimization of Gram matrix based cost in neural style transfer. We demonstrate the effectiveness of our algorithm by generating high-quality stylized images with comparisons to a number of recent methods. We also analyze our method by visualizing the whitened features and synthesizing textures via simple feature coloring.