Research map / Image generation / Text-to-image diffusion
VisualCloze: A Universal Image Generation Framework via Visual In-Context Learning
Zhong-Yu Li, Ruoyi Du, Juncheng Yan, Le Zhuo, Zhen Li, Peng Gao, Zhanyu Ma, Ming-Ming Cheng
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 |
|---|---|---|---|
| fine-tune FLUX.1-Fill-dev instead of FLUX.1-dev (ablation) | VisualCloze · Depth-to-image | RMSE: 25.06 -> 10.31 | Improved |
| fine-tune FLUX.1-Fill-dev instead of FLUX.1-dev (ablation) | VisualCloze · Canny-to-image | F1: 0.39 -> 0.35 | Hurt |
| fine-tune FLUX.1-Fill-dev instead of FLUX.1-dev (ablation) | VisualCloze · Subject-driven generation | DINOv2: 78.05 -> 80.41 | Improved |
| fine-tune FLUX.1-Fill-dev instead of FLUX.1-dev (ablation) | VisualCloze · Style transfer | CLIP text alignment: 0.30 -> 0.29 | No clear effect |
| add two in-context examples (ablation) | VisualCloze (Fill) · Depth-to-image | RMSE: 10.31 -> 9.68 | Improved |
| add two in-context examples (ablation) | VisualCloze (Fill) · Deblurring | RMSE: 26.53 -> 25.57 | Improved |
| add two in-context examples (ablation) | VisualCloze (Fill) · Canny-to-image | FID: 30.60 -> 31.15 | Hurt |
| add two in-context examples (ablation) | VisualCloze (Fill) · Subject-driven generation | DINOv2: 80.41 -> 80.32 | No clear effect |
| add one in-context example (ablation) | VisualCloze (Fill) · Deblurring | FID: 40.59 -> 36.93 | Improved |
| use VisualCloze (Fill) instead of OmniGen | VisualCloze · Style transfer | CLIP text alignment: 0.27 -> 0.29 | Improved |
| use VisualCloze (Fill) instead of OmniGen | VisualCloze · Canny-to-image | FID: 51.58 -> 30.60 | Improved |
| use VisualCloze (Fill) instead of OminiControl specialist | VisualCloze · Subject-driven generation | DINOv2: 73.17 -> 80.41 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.
- OmniGen — OmniGen: Unified Image Generation (CVPR 2025) · better on 5 of 6 shared results
- OminiControl — OminiControl: Minimal and Universal Control for Diffusion Transformer (ICCV 2025) · better on 1 of 3 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
- OmniGen: Unified Image Generation (CVPR 2025 · similarity 0.87)
- InstructCV: Instruction-Tuned Text-to-Image Diffusion Models as Vision Generalists (ICLR 2024 · similarity 0.87)
- Embracing Collaboration Over Competition: Condensing Multiple Prompts for Visual In-Context Learning (CVPR 2025 · similarity 0.87)
- Towards Universal Dataset Distillation via Task-Driven Diffusion (CVPR 2025 · similarity 0.86)
- VIGC: Visual Instruction Generation and Correction (AAAI 2024 · similarity 0.86)
- Instruct-Imagen: Image Generation with Multi-modal Instruction (CVPR 2024 · similarity 0.86)
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.
