Research map / Image generation / Super-resolution
Knowledge Distillation based Degradation Estimation for Blind Super-Resolution
Bin Xia, Yulun Zhang, Yitong Wang, Yapeng Tian, Wenming Yang, Radu Timofte, Luc Van Gool
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 |
|---|---|---|---|
| use the student IDR estimator with knowledge distillation rather than without it (ablation) | KDSR_S-M · Urban100, Gaussian8 | PSNR: 25.54 -> 25.96; 0.42dB | Improved |
| use IDR-DDC rather than ordinary convolution in IDR-DCRB (ablation) | KDSR_S-M · Urban100, Gaussian8 | PSNR: 25.37 -> 25.54; 0.17dB | Improved |
| estimate IDR rather than directly input oracle blur kernels (ablation) | KDSR_S-M · Urban100, Gaussian8 | PSNR: 26.08 -> 25.96 | No clear effect |
| distill from the teacher using KL loss rather than L1 loss (ablation) | KDSR_S-M · Urban100, Gaussian8 | PSNR: 25.92 -> 25.96 | Improved |
| distill from the teacher using KL loss rather than L2 loss (ablation) | KDSR_S-M · Urban100, Gaussian8 | PSNR: 25.88 -> 25.96 | Improved |
| use the student estimator instead of the paired-input teacher estimator (ablation) | KDSR · Urban100, Gaussian8 | PSNR: 26.16 -> 25.96; 0.2dB | Hurt |
| use KDSR_S-L rather than DCLS | Blind-SR · Set14, Gaussian8 | PSNR: 28.54 -> 28.68 | Improved |
| use KDSR_S-L rather than DCLS | Blind-SR · Set5, Gaussian8 | PSNR: 32.12 -> 32.11 | No clear effect |
| use KDSR_s-GAN rather than Real-ESRGAN | Real-world SR · AIM2019 | LPIPS: 0.3956 -> 0.3758 | Improved |
| use KDSR_s-GAN rather than Real-ESRGAN | Real-world SR · NTIRE2020 | PSNR: 26.40 -> 27.12 | Improved |
Closest papers by meaning
- LightBSR: Towards Lightweight Blind Super-Resolution via Discriminative Implicit Degradation Representation Learning (ICCV 2025 · similarity 0.93)
- Unsupervised Degradation Representation Aware Transform for Real-World Blind Image Super-Resolution (AAAI 2025 · similarity 0.92)
- Unsupervised Degradation Representation Learning for Blind Super-Resolution (CVPR 2021 · similarity 0.92)
- End-to-End Alternating Optimization for Real-World Blind Super Resolution (IJCV 2023 · similarity 0.91)
- Unsupervised Diffusion-Based Degradation Modeling for Real-World Super-Resolution (AAAI 2025 · similarity 0.88)
- Low-Res Leads the Way: Improving Generalization for Super-Resolution by Self-Supervised Learning (CVPR 2024 · similarity 0.88)
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.
