Research map / Image generation / Super-resolution
AdaDiffSR: Adaptive Region-aware Dynamic acceleration Diffusion Model for Real-World Image Super-Resolution
Yuanting Fan, Chengxu Liu, Nengzhong Yin, Changlong Gao, Xueming Qian
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 |
|---|---|---|---|
| increase skip-step intervals from (5,5,5,5) to (10,15,20,25) (ablation) | AdaDiffSR · DRealSR | LPIPS / runtime: LPIPS 0.2578 -> 0.2939, RT 13.4 -> 7.3 | Mixed |
| use both FR and NR IQA metrics in MMLE instead of only one type (ablation) | AdaDiffSR MMLE · RealSR | MUSIQ: 42.53 -> 60.47 (vs FR only) | Improved |
| use PFJ instead of concat feature fusion (ablation) | AdaDiffSR · DPED-iPhone | MUSIQ: 32.19 -> 51.84 | Improved |
| use PFJ instead of cross-attention fusion (ablation) | AdaDiffSR · DPED-iPhone | NIQE: 4.21 -> 3.09 | Improved |
| add DTSS with MMLE regressor instead of exact IQA metrics (ablation) | LDM baseline · DIV2K Valid 1024x1024 | runtime (s): 21.5 -> 18.1 | Improved |
| add progressive feature injection (PFJ) (ablation) | LDM + DTSS · DIV2K Valid 1024x1024 | PSNR: 22.38 -> 24.22 | Improved |
| add region integration strategy (RIS) (ablation) | AdaDiffSR · DIV2K Valid 1024x1024 | LPIPS: 0.2338 -> 0.2153 | No clear effect |
| use segmentation models (SAM etc.) for region slicing instead of fixed slicing (ablation) | AdaDiffSR · DIV2K Valid | runtime (s): 4.81 -> 8.59 (best alternative) | Hurt |
| apply AdaDiffSR dynamic timesteps vs StableSR | AdaDiffSR · complexity comparison | inference time and FLOPs: 1.5x fewer inference time, 2.7x fewer FLOPs | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- ResShift — ResShift: Efficient Diffusion Model for Image Super-resolution by Residual Shifting (NeurIPS 2023) · better on 1 of 2 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
- SinSR: Diffusion-Based Image Super-Resolution in a Single Step (CVPR 2024 · similarity 0.91)
- TSD-SR: One-Step Diffusion with Target Score Distillation for Real-World Image Super-Resolution (CVPR 2025 · similarity 0.90)
- Arbitrary-steps Image Super-resolution via Diffusion Inversion (CVPR 2025 · similarity 0.90)
- Taming Diffusion Prior for Image Super-Resolution with Domain Shift SDEs (NeurIPS 2024 · similarity 0.90)
- Unleashing the Power of One-Step Diffusion based Image Super-Resolution via a Large-Scale Diffusion Discriminator (NeurIPS 2025 · similarity 0.90)
- Binarized Diffusion Model for Image Super-Resolution (NeurIPS 2024 · similarity 0.90)
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.
