Research map / Video / Video segmentation and tracking
MuCAN: Multi-Correspondence Aggregation Network for Video Super-Resolution
Wenbo Li, Xin Tao, Taian Guo, Lu Qi, Jiangbo Lu, Jiaya Jia
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 TM-CAM temporal multi-correspondence aggregation (ablation) | Baseline VSR network · REDS test (lightweight setting) | PSNR: 28.98 -> 30.13 | Improved |
| add TM-CAM (ablation) | Baseline VSR network · REDS test (lightweight setting) | SSIM: 0.8280 -> 0.8614 | Improved |
| add CN-CAM cross-scale nonlocal aggregation (ablation) | Baseline+TM-CAM · REDS test (lightweight setting) | PSNR: 30.13 -> 30.25 | Improved |
| add edge-aware loss (EAL) (ablation) | Baseline+TM-CAM+CN-CAM · REDS test (lightweight setting) | PSNR: 30.25 -> 30.31 | Improved |
| add edge-aware loss (EAL) (ablation) | Baseline+TM-CAM+CN-CAM · REDS test (lightweight setting) | SSIM: 0.8641 -> 0.8648 | Improved |
| aggregate K=4 correspondences instead of 1 (ablation) | TM-CAM · REDS test (lightweight setting) | PSNR: 30.19 -> 30.31 | Improved |
| aggregate K=4 instead of 2 correspondences (ablation) | TM-CAM · REDS test (lightweight setting) | PSNR: 30.24 -> 30.31 | Improved |
| increase K from 4 to 6 (ablation) | TM-CAM · REDS test (lightweight setting) | PSNR: 30.31 -> 30.30 | No clear effect |
| use pixel-adaptive aggregation weights instead of fixed weights (ablation) | TM-CAM (K=4) · REDS test (lightweight setting) | PSNR: 30.12 -> 30.31 | Improved |
| use pixel-adaptive aggregation weights instead of fixed weights (ablation) | TM-CAM (K=4) · REDS test (lightweight setting) | SSIM: 0.8614 -> 0.8648 | Improved |
| use a larger patch size in TM-CAM (ablation) | TM-CAM · REDS test (lightweight setting) | PSNR | Improved |
| use MuCAN instead of EDVR (trained without pretraining) | VSR network · REDS test (4 clips, x4) | PSNR: 30.71 -> 30.88 | Improved |
Later papers that compared with this method
1 later accepted paper reports MuCAN as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- COMISR: Compression-Informed Video Super-Resolution (ICCV 2021 · 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
- BasicVSR: The Search for Essential Components in Video Super-Resolution and Beyond (CVPR 2021 · similarity 0.88)
- Sequence Matters: Harnessing Video Models in 3D Super-Resolution (AAAI 2025 · similarity 0.88)
- Rethinking Alignment in Video Super-Resolution Transformers (NeurIPS 2022 · similarity 0.88)
- SAVSR: Arbitrary-Scale Video Super-Resolution via a Learned Scale-Adaptive Network (AAAI 2024 · similarity 0.86)
- SeeClear: Semantic Distillation Enhances Pixel Condensation for Video Super-Resolution (NeurIPS 2024 · similarity 0.86)
- One-Step Diffusion for Detail-Rich and Temporally Consistent Video Super-Resolution (NeurIPS 2025 · similarity 0.86)
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.
