Research map / Image generation / Super-resolution
Consistency Trajectory Matching for One-Step Generative Super-Resolution
Weiyi You, Mingyang Zhang, Leheng Zhang, Xingyu Zhou, Kexuan Shi, Shuhang Gu
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 Distribution Trajectory Matching loss to consistency training (ablation) | CTMSR · ImageNet-Test | CLIPIQA: 0.6092 -> 0.6913 | Improved |
| add Distribution Trajectory Matching loss to consistency training (ablation) | CTMSR · ImageNet-Test | MUSIQ: 56.650 -> 60.142 | Improved |
| add DTM loss (ablation) | CTMSR · ImageNet-Test | PSNR: 24.71 -> 24.73 | Improved |
| use score distillation sampling loss instead of DTM (ablation) | CTMSR · ImageNet-Test | PSNR: 24.71 -> 23.17 | Hurt |
| use SDS instead of consistency training alone (ablation) | CTMSR · ImageNet-Test | MUSIQ: 56.650 -> 58.188 | Improved |
| replace SDS by DTM (ablation) | CTMSR · ImageNet-Test | MUSIQ: 58.188 -> 60.142 | Improved |
| one-step distillation-free consistency training instead of distilling ResShift (SinSR) | one-step SR · ImageNet-Test | MUSIQ: 53.632 -> 60.142 | Improved |
| one-step CTMSR versus SinSR on real data | RealSR · RealSR | CLIPIQA: 0.6627 -> 0.6449 | Hurt |
| one-step CTMSR versus ResShift-4 on fidelity | ImageNet-Test · ImageNet-Test | PSNR: 25.02 -> 24.73 | Hurt |
| reduce inference to one step (CTMSR-1 versus ResShift-15) | inference runtime · 64x64 input RTX3090 | runtime (ms): 689 -> 48 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- SinSR-1 — SinSR: Diffusion-Based Image Super-Resolution in a Single Step (CVPR 2024) · better on 7 of 8 shared results
- ResShift-4 — ResShift: Efficient Diffusion Model for Image Super-resolution by Residual Shifting (NeurIPS 2023) · better on 2 of 3 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
- SinSR: Diffusion-Based Image Super-Resolution in a Single Step (CVPR 2024 · similarity 0.91)
- One Diffusion Step to Real-World Super-Resolution via Flow Trajectory Distillation (ICML 2025 · similarity 0.91)
- TSD-SR: One-Step Diffusion with Target Score Distillation for Real-World Image Super-Resolution (CVPR 2025 · similarity 0.90)
- Taming Diffusion Prior for Image Super-Resolution with Domain Shift SDEs (NeurIPS 2024 · similarity 0.89)
- Noise-Started One-Step Image Super-Resolution via LR-Conditioned SplitMeanFlow and GAN Refinement (NeurIPS 2026 · similarity 0.89)
- Solving Diffusion ODEs with Optimal Boundary Conditions for Better Image Super-Resolution (ICLR 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.
