Research map / Image generation / Super-resolution
PIPAL: a Large-Scale Image Quality Assessment Dataset for Perceptual Image Restoration
Jinjin Gu, Haoming Cai, Haoyu Chen, Xiaoxing Ye, Jimmy S. Ren, Chao Dong
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 |
|---|---|---|---|
| replace max pooling with BlurPool (ablation) | LPIPS-Alex · PIPAL (full set) | SRCC: 0.5604 -> 0.5918 | Improved |
| replace max pooling with BlurPool (ablation) | LPIPS-Alex · PIPAL (GAN-based distortion subset) | SRCC: 0.4862 -> 0.5135 | Improved |
| replace max pooling with l2 pooling (ablation) | LPIPS-Alex · PIPAL (full set) | SRCC: 0.5604 -> 0.5816 | Improved |
| replace max pooling with l2 pooling (ablation) | LPIPS-Alex · PIPAL (GAN-based distortion subset) | SRCC: 0.4862 -> 0.4942 | Improved |
| evaluate on GAN-based SR outputs instead of PSNR-oriented SR | PieAPP · PIPAL | SRCC: 0.8097 -> 0.5530 | Hurt |
| evaluate on GAN-based SR outputs instead of PSNR-oriented SR | IFC · PIPAL | SRCC: 0.8244 -> 0.3217 | Hurt |
| use GAN-based SR (SRGAN) instead of EDSR | SR model · PIPAL x4 SR benchmark | MOS (Elo): 1447.44 -> 1494.14 | Improved |
| use GAN-based SR (SRGAN) instead of EDSR | SR model · PIPAL x4 SR benchmark | PSNR (dB): 25.17 -> 22.57 | Hurt |
| optimize PI further (RankSRGAN vs ESRGAN) | SR model · PIPAL x4 SR benchmark | MOS (Elo): 1534.25 -> 1518.29 | Hurt |
| optimize PI further (RankSRGAN vs ESRGAN) | SR model · PIPAL x4 SR benchmark | PI: 3.2198 -> 2.5636 | Improved |
| FSIM_C on spatial warping vs GAN-based distortion | FSIMc · PIPAL | SRCC: 0.31 vs 0.41 | No clear effect |
| evaluate PieAPP on PIPAL instead of TID2013 | PieAPP · IQA benchmark | SRCC: 0.90 -> 0.71 | Hurt |
Closest papers by meaning
- UARE: A Unified Vision-Language Model for Image Quality Assessment, Restoration, and Enhancement (CVPR 2026 · similarity 0.88)
- Augmenting Perceptual Super-Resolution via Image Quality Predictors (CVPR 2025 · similarity 0.88)
- Adaptive Image Quality Assessment via Teaching Large Multimodal Model to Compare (NeurIPS 2024 · similarity 0.88)
- Toward Generalized Image Quality Assessment: Relaxing the Perfect Reference Quality Assumption (CVPR 2025 · similarity 0.88)
- Causal-IQA: Towards the Generalization of Image Quality Assessment Based on Causal Inference (ICML 2024 · similarity 0.87)
- Boosting Image Quality Assessment through Efficient Transformer Adaptation with Local Feature Enhancement (CVPR 2024 · 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.
