# OmniStyle2: Learning to Stylize by Learning to Destylize

arXiv: https://arxiv.org/abs/2509.05970 · HTML: https://mlsanity.com/paper/2509.05970/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 35 of its 44 references that are in the map.
- First posted on arXiv 7 Sep 2025.

## Builds on (most cited first, 30 of 35)

- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023)
- [Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization](https://mlsanity.com/paper/1703.06868/) (2017)
- [Qwen-Image Technical Report](https://mlsanity.com/paper/2508.02324/) (2025)
- [A Neural Algorithm of Artistic Style](https://mlsanity.com/paper/1508.06576/) (2015)
- [Universal Style Transfer via Feature Transforms](https://mlsanity.com/paper/1705.08086/) (2017)
- [Inversion-Based Style Transfer with Diffusion Models](https://mlsanity.com/paper/2211.13203/) (2022)
- [P+: Extended Textual Conditioning in Text-to-Image Generation](https://mlsanity.com/paper/2303.09522/) (2023)
- [StyleDrop: Text-to-Image Generation in Any Style](https://mlsanity.com/paper/2306.00983/) (2023)
- [Style Injection in Diffusion: A Training-free Approach for Adapting Large-scale Diffusion Models for Style Transfer](https://mlsanity.com/paper/2312.09008/) (2023)
- [ZipLoRA: Any Subject in Any Style by Effectively Merging LoRAs](https://mlsanity.com/paper/2311.13600/) (2023)
- [InstantStyle: Free Lunch towards Style-Preserving in Text-to-Image Generation](https://mlsanity.com/paper/2404.02733/) (2024)
- [Style Transfer by Relaxed Optimal Transport and Self-Similarity](https://mlsanity.com/paper/1904.12785/) (2019)
- [StyleDiffusion: Controllable Disentangled Style Transfer via Diffusion Models](https://mlsanity.com/paper/2308.07863/) (2023)
- [Implicit Style-Content Separation using B-LoRA](https://mlsanity.com/paper/2403.14572/) (2024)
- [Visual Attribute Transfer through Deep Image Analogy](https://mlsanity.com/paper/1705.01088/) (2017)
- [Exact Feature Distribution Matching for Arbitrary Style Transfer and Domain Generalization](https://mlsanity.com/paper/2203.07740/) (2022)
- [Measuring Style Similarity in Diffusion Models](https://mlsanity.com/paper/2404.01292/) (2024)
- [Controlling Perceptual Factors in Neural Style Transfer](https://mlsanity.com/paper/1611.07865/) (2016)
- [Domain Enhanced Arbitrary Image Style Transfer via Contrastive Learning](https://mlsanity.com/paper/2205.09542/) (2022)
- [CSGO: Content-Style Composition in Text-to-Image Generation](https://mlsanity.com/paper/2408.16766/) (2024)
- [Arbitrary Style Transfer via Multi-Adaptation Network](https://mlsanity.com/paper/2005.13219/) (2020)
- [DEADiff: An Efficient Stylization Diffusion Model with Disentangled Representations](https://mlsanity.com/paper/2403.06951/) (2024)
- [StyleShot: A Snapshot on Any Style](https://mlsanity.com/paper/2407.01414/) (2024)
- [InstantStyle-Plus: Style Transfer with Content-Preserving in Text-to-Image Generation](https://mlsanity.com/paper/2407.00788/) (2024)
- [K-LoRA: Unlocking Training-Free Fusion of Any Subject and Style LoRAs](https://mlsanity.com/paper/2502.18461/) (2025)
- [AesPA-Net: Aesthetic Pattern-Aware Style Transfer Networks](https://mlsanity.com/paper/2307.09724/) (2023)
- [Attention Distillation: A Unified Approach to Visual Characteristics Transfer](https://mlsanity.com/paper/2502.20235/) (2025)
- [Multimodal Style Transfer via Graph Cuts](https://mlsanity.com/paper/1904.04443/) (2019)
- [StyleTokenizer: Defining Image Style by a Single Instance for Controlling Diffusion Models](https://mlsanity.com/paper/2409.02543/) (2024)

## Built on by (most cited first, 1 of 1)

- [Mixture of Style Experts for Diverse Image Stylization](https://mlsanity.com/paper/2603.16649/) (2026)

## Abstract (arXiv)

This paper introduces a scalable paradigm for supervised style transfer by inverting the problem: instead of learning to stylize directly, we learn to destylize, reducing stylistic elements from artistic images to recover their natural counterparts and thereby producing authentic, pixel-aligned training pairs at scale. To realize this paradigm, we propose DeStylePipe, a progressive, multi-stage destylization framework that begins with global general destylization, advances to category-wise instruction adaptation, and ultimately deploys specialized model adaptation for complex styles that prompt engineering alone cannot handle. Tightly integrated into this pipeline, DestyleCoT-Filter employs Chain-of-Thought reasoning to assess content preservation and style removal at each stage, routing challenging samples forward while discarding persistently low-quality pairs. Built on this framework, we construct DeStyle-350K, a large-scale dataset aligning diverse artistic styles with their underlying content. We further introduce BCS-Bench, a benchmark featuring balanced content generality and style diversity for systematic evaluation. Extensive experiments demonstrate that models trained on DeStyle-350K achieve superior stylization quality, validating destylization as a reliable and scalable supervision paradigm for style transfer.