Research map / Image generation / Text-to-image diffusion
In-Context Learning Unlocked for Diffusion Models
Zhendong Wang, Yifan Jiang, Yadong Lu, yelong shen, Pengcheng He, Weizhu Chen, Zhangyang Wang, Mingyuan Zhou
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 |
|---|---|---|---|
| joint multi-task in-context training instead of per-task ControlNet finetuning | Prompt Diffusion · Depth-to-Image | FID: 19.81 -> 18.60 | Improved |
| joint multi-task in-context training instead of per-task ControlNet finetuning | Prompt Diffusion · Hed-to-Image | FID: 13.07 -> 13.35 | No clear effect |
| joint multi-task in-context training instead of per-task ControlNet finetuning | Prompt Diffusion · Seg-to-Image | FID: 20.71 -> 19.46 | Improved |
| joint multi-task in-context training instead of per-task ControlNet finetuning | Prompt Diffusion · Image-to-Depth | RMSE: 0.20 -> 0.21 | No clear effect |
| joint multi-task in-context training instead of per-task ControlNet finetuning | Prompt Diffusion · Image-to-Hed | RMSE: 0.18 -> 0.14 | Improved |
| joint multi-task in-context training instead of per-task ControlNet finetuning | Prompt Diffusion · Image-to-Seg | RMSE: 0.36 -> 0.31 | Improved |
| apply a task-specific finetuned ControlNet to unseen tasks | CN(FT) · unseen tasks (qualitative) | generation quality | Hurt |
| use different random images to build the example pair (ablation) | Prompt Diffusion · qualitative ablation | output | No clear effect |
| misalign type of query image with example source image (ablation) | Prompt Diffusion · qualitative ablation | output predictability | Hurt |
Closest papers by meaning
- InstructCV: Instruction-Tuned Text-to-Image Diffusion Models as Vision Generalists (ICLR 2024 · similarity 0.91)
- Prompting Hard or Hardly Prompting: Prompt Inversion for Text-to-Image Diffusion Models (CVPR 2024 · similarity 0.91)
- Diff-Prompt: Diffusion-Driven Prompt Generator with Mask Supervision (ICLR 2025 · similarity 0.90)
- Dual Diffusion for Unified Image Generation and Understanding (CVPR 2025 · similarity 0.90)
- AP-Adapter: Improving Generalization of Automatic Prompts on Unseen Text-to-Image Diffusion Models (NeurIPS 2024 · similarity 0.90)
- PromptFix: You Prompt and We Fix the Photo (NeurIPS 2024 · similarity 0.89)
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.
