Research map / Image generation / Super-resolution
PS-SR: Pseudo-Single-Step Video Super-Resolution via Speculative Diffusion
Aiqiu Wu, Zhaofan Qiu, Ting Yao, Tao Mei
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 frequency-domain update rule (high-pass constrained refinement) (ablation) | PS-SR · SPMCS | PSNR: 18.661 -> 22.092 | Improved |
| add frequency-domain update rule (ablation) | PS-SR · SPMCS | MUSIQ: 67.066 -> 61.004 | Hurt |
| add VSD loss to base model (ablation) | PS-SR · SPMCS | MUSIQ: 56.369 -> 61.004 | Improved |
| add adversarial loss (ablation) | PS-SR · SPMCS | MUSIQ: 56.573 -> 61.004 | Improved |
| add pixel-space loss (ablation) | PS-SR · SPMCS | MUSIQ: 56.820 -> 61.004 | Improved |
| add pixel-space loss (ablation) | PS-SR · SPMCS | PSNR: 22.266 -> 22.092 | No clear effect |
| increase speculative steps T from 1 to 4 (ablation) | PS-SR · SPMCS | MUSIQ: 50.632 -> 61.004 | Improved |
| increase speculative steps T from 1 to 4 (ablation) | PS-SR · SPMCS | PSNR: 22.337 -> 22.092 | Hurt |
| increase refinement strength alpha from 0.2 to 1.0 (ablation) | PS-SR · SPMCS | PSNR: 22.272 -> 21.774 | Hurt |
| increase refinement strength alpha from 0.2 to 1.0 (ablation) | PS-SR · SPMCS | MUSIQ: 53.899 -> 66.535 | Improved |
| prune 25 of 30 draft blocks instead of 20 (ablation) | PS-SR · SPMCS | MUSIQ: 61.004 -> 56.936 | Hurt |
| prune 20 of 30 draft blocks instead of 0 (ablation) | PS-SR · SPMCS | MUSIQ: 61.484 -> 61.004 | No clear effect |
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.
- DOVE — DOVE: Efficient One-Step Diffusion Model for Real-World Video Super-Resolution (NeurIPS 2025) · better on 8 of 10 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
- SkipSR: Faster Super-Resolution with Token Skipping (NeurIPS 2026 · similarity 0.92)
- DOVE: Efficient One-Step Diffusion Model for Real-World Video Super-Resolution (NeurIPS 2025 · similarity 0.91)
- SinSR: Diffusion-Based Image Super-Resolution in a Single Step (CVPR 2024 · similarity 0.91)
- One-Step Diffusion for Detail-Rich and Temporally Consistent Video Super-Resolution (NeurIPS 2025 · similarity 0.90)
- SeeClear: Semantic Distillation Enhances Pixel Condensation for Video Super-Resolution (NeurIPS 2024 · similarity 0.90)
- PatchVSR: Breaking Video Diffusion Resolution Limits with Patch-wise Video Super-Resolution (CVPR 2025 · 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.
