Research map / Image generation / Super-resolution
Evaluating the Generalization Ability of Super-Resolution Networks
Yihao Liu, Hengyuan Zhao, Jinjin Gu, Yu Qiao, 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 |
|---|---|---|---|
| train with blur0-4 degradation data (ablation) | SRResNet · PIES-Blur | mSRGA: 3.639 -> 2.967 | Improved |
| train with noise0-20 data (ablation) | SRResNet · PIES-Noise | mSRGA: 3.727 -> 2.325 | Improved |
| train with noise0-20 data (ablation) | SRResNet · PIES-Blur | mSRGA: 3.639 -> 3.623 | No clear effect |
| train with both blur0-4 and noise0-20 (ablation) | SRResNet · PIES-Noise | mSRGA: 3.727 -> 1.574 | Improved |
| train with both blur0-4 and noise0-20 (ablation) | SRResNet · PIES-RealCam | mSRGA: 3.563 -> 2.566 | Improved |
| train with both blur0-4 and noise0-20 (ablation) | SRResNet · PIES-RealLQ | mSRGA: 3.825 -> 3.736 | No clear effect |
| use SwinIR-GAN instead of Real-ESRGAN (ablation) | blind SR model · PIES-Blur | mSRGA: 2.480 -> 1.639 | Improved |
| use SwinIR-PSNR instead of Real-ESRNet-style GAN models (ablation) | blind SR model · PIES-BlurNoise | mSRGA: 2.692 -> 2.297 | Improved |
| use training with more synthetic degradations (Real-ESRGAN) instead of DASR (ablation) | blind SR model · PIES-RealLQ | mSRGA: 4.033 -> 3.823 | Improved |
| use DAN instead of IKC (ablation) | blind SR model · PIES-RealCam | mSRGA: 3.432 -> 3.937 | Hurt |
| use SwinIR-GAN under severe noise (ablation) | SwinIR-GAN · PIES-Noise | SRGA | Mixed |
| use SRGA instead of PSNR under luminance jitter in [-10, +20] | evaluation metric · PIES-Blur | SRGA stability | Improved |
Closest papers by meaning
- Navigating Beyond Dropout: An Intriguing Solution towards Generalizable Image Super Resolution (CVPR 2024 · similarity 0.88)
- Image Processing GNN: Breaking Rigidity in Super-Resolution (CVPR 2024 · similarity 0.87)
- Hitchhiker's Guide to Super-Resolution: Introduction and Recent Advances (TPAMI 2023 · similarity 0.86)
- An In-depth Investigation of Sparse Rate Reduction in Transformer-like Models (NeurIPS 2024 · similarity 0.85)
- Recursive Generalization Transformer for Image Super-Resolution (ICLR 2024 · similarity 0.85)
- GSDD: Generative Space Dataset Distillation for Image Super-resolution (AAAI 2024 · similarity 0.85)
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.
