Research map / Video / Video segmentation and tracking
Self-Supervised Deep Blind Video Super-Resolution
Haoran Bai, Jinshan Pan
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 both self-supervised constraints (ablation) | self-supervised blind video SR · REDS4 | PSNR: 25.73 -> 29.23 | Improved |
| add kernel-related constraint (ablation) | self-supervised blind video SR · REDS4 | PSNR: 27.66 -> 29.23 | Improved |
| add auxiliary supervised constraint (ablation) | self-supervised blind video SR · SPMCS | PSNR: 25.67 -> 27.77 | Improved |
| add detach so the constraint only updates kernel network (ablation) | self-supervised blind video SR · VID4 | PSNR: 23.03 -> 24.59 | Improved |
| add temporal information (warped auxiliary frames) to the constraint (ablation) | self-supervised blind video SR · REDS4 | PSNR: 28.40 -> 29.23 | Improved |
| add temporal information (ablation) | self-supervised blind video SR · REDS4 | SSIM: 0.8248 -> 0.8453 | Improved |
| add both constraints, effect on kernel quality (ablation) | self-supervised blind video SR · REDS4 regenerated LR | PSNR: 30.49 -> 39.62 | Improved |
| online self-supervised fine-tuning on real videos (ablation) | self-supervised blind video SR · real videos | NIQE: 13.357 -> 12.483 | Improved |
| use HR videos as supervision instead of self-supervision (ablation) | self-supervised blind video SR · SPMCS | PSNR: 27.77 -> 27.99 | No clear effect |
| self-supervised method vs supervised EDVR on Gaussian-blur REDS4 | EDVR · REDS4 | PSNR: 27.91 -> 29.23 | Improved |
| add explicit blur kernel estimation vs blind image SR methods | IKC · benchmarks | PSNR: 1.06dB | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TDAN — TDAN: Temporally-Deformable Alignment Network for Video Super-Resolution (CVPR 2020) · better on 4 of 4 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
- Deep Blind Video Super-Resolution (ICCV 2021 · similarity 0.91)
- Spectrum-to-Kernel Translation for Accurate Blind Image Super-Resolution (NeurIPS 2021 · similarity 0.89)
- Low-Res Leads the Way: Improving Generalization for Super-Resolution by Self-Supervised Learning (CVPR 2024 · similarity 0.89)
- Blind Super-Resolution With Iterative Kernel Correction (CVPR 2019 · similarity 0.89)
- Learning the Non-Differentiable Optimization for Blind Super-Resolution (CVPR 2021 · similarity 0.88)
- Self-supervised ControlNet with Spatio-Temporal Mamba for Real-world Video Super-resolution (CVPR 2025 · 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.
