Research map / Image generation / Autoregressive and latent generation
HiPrompt: Tuning-free Higher-Resolution Generation with Hierarchical MLLM Prompts
Xinyu Liu, Yingqing He, Lanqing Guo, Xiang Li, Bu Jin, Yan Li, Chi-Min Chan, Wei Xue, Wenhan Luo, Qifeng Liu, Yike Guo
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 hierarchical MLLM prompt guidance (ablation) | SDXL-based DemoFusion baseline · 2048x4096 | FID_r: 1.21 decrease | Improved |
| add noise decomposition on top of MLLM prompts (ablation) | HiPrompt · 2048x4096 | FID_c: 46.93 -> 35.17 | Improved |
| add both MLLM prompts and noise decomposition (ablation) | baseline · 2048x4096 | FID_r: 76.24 -> 73.82 | Improved |
| add LLaVA prompts only (ablation) | baseline · 4096x4096 | FID_c: 14.32 -> 11.82 | Improved |
| add ShareCaptioner prompts only (ablation) | baseline · 4096x4096 | FID_c: 14.32 -> 8.80 | Improved |
| add noise decomposition to LLaVA prompts (ablation) | HiPrompt · 4096x4096 | FID_r: 71.81 -> 70.40 | Improved |
| add noise decomposition to ShareCaptioner prompts (ablation) | HiPrompt · 4096x4096 | FID_r: 71.01 -> 70.22 | Improved |
| swap LLaVA for ShareCaptioner as MLLM (ablation) | HiPrompt · 4096x4096 | FID_c: 11.82 -> 8.80 | Improved |
| add N-grams refinement (ablation) | HiPrompt with LLaVA · 2048x2048 | FID_r: 68.78 -> 67.79 | Improved |
| add LLaVA prompts only (no ND) (ablation) | baseline · 4096x4096 | FID_r: 71.04 -> 71.81 | Hurt |
| HiPrompt vs DemoFusion at 16x | HiPrompt · 4096x4096 | FID_c: 36.89 -> 43.60 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- DemoFusion — DemoFusion: Democratising High-Resolution Image Generation With No $$$ (CVPR 2024) · better on 5 of 6 shared results
- ScaleCrafter — ScaleCrafter: Tuning-free Higher-Resolution Visual Generation with Diffusion Models (ICLR 2024) · better on 6 of 6 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
- ResMaster: Mastering High-Resolution Image Generation via Structural and Fine-Grained Guidance (AAAI 2025 · similarity 0.90)
- ScaleCrafter: Tuning-free Higher-Resolution Visual Generation with Diffusion Models (ICLR 2024 · similarity 0.89)
- DiffuseHigh: Training-Free Progressive High-Resolution Image Synthesis Through Structure Guidance (AAAI 2025 · similarity 0.88)
- UltraPixel: Advancing Ultra High-Resolution Image Synthesis to New Peaks (NeurIPS 2024 · similarity 0.87)
- FreeControl: Efficient, Training-Free Structural Control via One-Step Attention Extraction (NeurIPS 2025 · similarity 0.87)
- RaPD: Resolution-Agnostic Pixel Diffusion via Semantics-Enriched Implicit Representations (NeurIPS 2026 · similarity 0.87)
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.
