SplitFlux: Learning to Decouple Content and Style from a Single Image

Yitong Yang¹,   Yinglin Wang^1†,   Changshuo Wang³,   Yongjun Zhang⁴,   Ziyang Chen⁴,   Shuting He^1,2†,  

¹School of Computing and Artificial Intelligence, Shanghai University of Finance and Economics, Shanghai, China.
²MoE Key Laboratory of Interdisciplinary Research of Computation and Economics, Shanghai, China.
³Department of Computer Science University College London, London, United Kingdom.
⁴College of Computer Science and Technology, Guizhou University
^†Corresponding Author.

   

📢 News

• [2026-02] Our paper is accepted by CVPR 2026! 🎉
• [2026-03] We have open-sourced the code.

💡 Overview

We conduct a systematic analysis of Flux and make two key observations: (1) Single Stream Blocks are essential for image generation; and (2) Early single stream blocks mainly control content, whereas later blocks govern style. Based on these insights, we propose SplitFlux, which disentangles content and style by fine-tuning the single stream blocks via LoRA, enabling the disentangled content to be re-embedded into new contexts. It includes two key components: (1) Rank-Constrained Adaptation. To preserve content identity and structure, we compress the rank and amplify the magnitude of updates within specific blocks, preventing content leakage into style blocks. (2) Visual-Gated LoRA. We split the content LoRA into two branches with different ranks, guided by image saliency. The high-rank branch preserves primary subject information, while the low-rank branch encodes residual details, mitigating content overfitting and enabling seamless re-embedding.

🔧 Installation

git clone https://github.com/yangyt46/SplitFlux.git
cd SplitFlux
conda create -n splitflux python=3.10
conda activate splitflux
pip install -r requirements.txt

🚀 Run

Training

bash run.sh

Note: You can replace “object” with the specific object in the image to achieve better results in Recontextualization.

Inference

For disentanglement:

# content
python infer_content.py
#style
python infer_style.py

For merger:

python infer_merger.py

For recontextualization:

# content
python infer_recontext_content.py
# image
python infer_recontext_image.py

🎓 Citing SplitFlux

If you use SplitFlux in your research, please use the following BibTeX entry.

@article{yang2025splitflux,
  title={SplitFlux: Learning to Decouple Content and Style from a Single Image},
  author={Yang, Yitong and Wang, Yinglin and Wang, Changshuo and Zhang, Yongjun and Chen, Ziyang and He, Shuting},
  journal={arXiv preprint arXiv:2511.15258},
  year={2025}
}