Research map / Image generation / Editing and style transfer
AnyStyle: A Single LoRA is Sufficient for Image-Guided Style Transfer
Yongwen Lai, Chaoqun Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add the single style LoRA (ablation) | AnyStyle · Style transfer (40 images) | CLIP-Style: 57.3314 -> 68.2336 | Improved |
| add the single style LoRA (ablation) | AnyStyle · Style transfer (40 images) | CLIP-T: 29.9431 -> 32.0717 | Improved |
| add the single style LoRA (ablation) | AnyStyle · Style transfer (40 images) | DINO-Style: 0.6170 -> 0.7234 | Improved |
| modulate query only instead of key only (ablation) | AnyStyle · Style transfer (40 images) | DS-Content: 0.5198 -> 0.4210 | Improved |
| modulate query only instead of key only (ablation) | AnyStyle · Style transfer (40 images) | DINO-Style: 0.7225 -> 0.7147 | Hurt |
| modulate query only instead of query+key (ablation) | AnyStyle · Style transfer (40 images) | CLIP-Style: 67.1565 -> 67.9291 | Improved |
| use single style LoRA instead of dual-LoRA | AnyStyle · Style transfer (40 images) | CLIP-Style: 63.3469 -> 67.9291 | Improved |
| use our style LoRA instead of B-LoRA's | AnyStyle · Style transfer (40 images) | DINO-Style: 0.7070 -> 0.7234 | Improved |
| use our style LoRA instead of UnZipLoRA's | AnyStyle · Style transfer (40 images) | DINO-Style: 0.7153 -> 0.7234 | Improved |
| use attention-based content injection instead of LoRA-based | content reconstruction · Reconstruction | PSNR: 11.7899 -> 18.0029 | Improved |
| swap FLUX backbone for SDXL | AnyStyle · Style transfer (40 images) | CLIP-Style: 67.9291 -> 66.4223 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- UnZipLoRA — UnZipLoRA: Separating Content and Style from a Single Image (ICCV 2025) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- OmniStyle: Filtering High Quality Style Transfer Data at Scale (CVPR 2025 · similarity 0.89)
- SigStyle: Signature Style Transfer via Personalized Text-to-Image Models (AAAI 2025 · similarity 0.89)
- StyleSSP: Sampling StartPoint Enhancement for Training-free Diffusion-based Method for Style Transfer (CVPR 2025 · similarity 0.89)
- Style Injection in Diffusion: A Training-free Approach for Adapting Large-scale Diffusion Models for Style Transfer (CVPR 2024 · similarity 0.88)
- Z*: Zero-shot Style Transfer via Attention Reweighting (CVPR 2024 · similarity 0.88)
- CLeaR: A Unified Framework for Resolving the Leakage–Degradation Dilemma in Style Transfer (NeurIPS 2026 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
