StyleBrush: Style Extraction and Transfer from a Single Image

TL;DR

StyleBrush is a novel method for style transfer from a single reference image that effectively separates style from structure, enabling high-quality stylization with a new dataset and achieving state-of-the-art results.

Contribution

We introduce StyleBrush, a new architecture with dual branches for style extraction and structural guidance, and create a large high-quality style image dataset for training.

Findings

Achieves state-of-the-art stylization quality

Uses a new dataset of 100K style images

Demonstrates effective style-structure separation

Abstract

Stylization for visual content aims to add specific style patterns at the pixel level while preserving the original structural features. Compared with using predefined styles, stylization guided by reference style images is more challenging, where the main difficulty is to effectively separate style from structural elements. In this paper, we propose StyleBrush, a method that accurately captures styles from a reference image and ``brushes'' the extracted style onto other input visual content. Specifically, our architecture consists of two branches: ReferenceNet, which extracts style from the reference image, and Structure Guider, which extracts structural features from the input image, thus enabling image-guided stylization. We utilize LLM and T2I models to create a dataset comprising 100K high-quality style images, encompassing a diverse range of styles and contents with high aesthetic score. To construct training pairs, we crop different regions of the same training image. Experiments show that our approach achieves state-of-the-art results through both qualitative and quantitative analyses. We will release our code and dataset upon acceptance of the paper.