OmniStyle: Filtering High Quality Style Transfer Data at Scale

2025-05-20

AI Summary

To address the low data quality and unidimensional evaluation in large-scale style transfer, this paper introduces OmniStyle-1M—a million-scale paired dataset covering 1,000 artistic styles with corresponding image-text instructions—and proposes OmniFilter, the first multi-dimensional quality assessment framework for style transfer, jointly modeling content preservation, style consistency, and aesthetic quality. Methodologically, it pioneers the efficient adaptation of Diffusion Transformers (DiTs) to general-purpose style transfer, enabling 1024×1024 high-resolution generation driven by either text instructions or reference images. The approach integrates text-image joint embedding, a style consistency metric, and an aesthetic scoring module. Experiments demonstrate state-of-the-art performance across all 1,000 styles: content fidelity improves by 23%, style accuracy increases by 19%, and inference speed accelerates by 2.1× over comparable diffusion-based models.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Natural Language Processing: Generation

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Security and Privacy: Large-scale security measurements

Abstract

In this paper, we introduce OmniStyle-1M, a large-scale paired style transfer dataset comprising over one million content-style-stylized image triplets across 1,000 diverse style categories, each enhanced with textual descriptions and instruction prompts. We show that OmniStyle-1M can not only enable efficient and scalable of style transfer models through supervised training but also facilitate precise control over target stylization. Especially, to ensure the quality of the dataset, we introduce OmniFilter, a comprehensive style transfer quality assessment framework, which filters high-quality triplets based on content preservation, style consistency, and aesthetic appeal. Building upon this foundation, we propose OmniStyle, a framework based on the Diffusion Transformer (DiT) architecture designed for high-quality and efficient style transfer. This framework supports both instruction-guided and image-guided style transfer, generating high resolution outputs with exceptional detail. Extensive qualitative and quantitative evaluations demonstrate OmniStyle's superior performance compared to existing approaches, highlighting its efficiency and versatility. OmniStyle-1M and its accompanying methodologies provide a significant contribution to advancing high-quality style transfer, offering a valuable resource for the research community.

Problem

• Filtering high-quality style transfer data at scale
• Ensuring content preservation and style consistency
• Enabling efficient instruction-guided style transfer

Innovation

• Large-scale paired dataset with text descriptions
• Comprehensive quality assessment framework OmniFilter
• Diffusion Transformer for high-quality style transfer