Research map / Video / Video segmentation and tracking
FMA-Net++: Motion- and Exposure-Aware Joint Video Super-Resolution and Deblurring
Geunhyuk Youk, Jihyong Oh, Munchurl Kim
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 hierarchical HRBA instead of sliding-window temporal modeling (ablation) | FMA-Net++ · REDS4-ME-5:5 (x4 VSRDB) | PSNR: 28.57 -> 29.24 | Improved |
| use hierarchical HRBA instead of recurrent propagation (ablation) | FMA-Net++ · REDS4-ME-5:5 (x4 VSRDB) | PSNR: 29.11 -> 29.24 | Improved |
| add ETE exposure conditioning (vs capacity-matched baseline) (ablation) | FMA-Net++ · REDS-RE (x4 VSRDB) | PSNR: 29.88 -> 30.13 | Improved |
| add ETE exposure conditioning (vs capacity-matched baseline) (ablation) | FMA-Net++ · REDS4-ME-5:4 | PSNR: 29.67 -> 29.66 | No clear effect |
| add ETE exposure conditioning (vs capacity-matched baseline) (ablation) | FMA-Net++ · GoPro (x4 VSRDB, unseen) | PSNR: 29.85 -> 30.49 | Improved |
| use contrastive ETE instead of frame-difference conditioning (ablation) | FMA-Net++ · REDS-RE (x4 VSRDB) | PSNR: 29.75 -> 30.13 | Improved |
| use contrastive ETE instead of ordinal-regression ETE (ablation) | FMA-Net++ · REDS-RE (x4 VSRDB) | PSNR: 29.97 -> 30.13 | Improved |
| provide correct guidance instead of random wrong-level guidance (ablation) | FMA-Net++ · REDS-RE (x4 VSRDB) | PSNR: 29.47 -> 30.13 | Improved |
| increase HRBA blocks from 1 to 4 (ablation) | FMA-Net++ · REDS4-ME-5:5 (x4 VSRDB) | PSNR: 28.29 -> 29.24 | Improved |
| use DA attention instead of SFT modulation (ablation) | FMA-Net++ · REDS4-ME-5:5 (x4 VSRDB) | PSNR: 28.86 -> 29.24 | Improved |
| increase multi-flow-mask pairs from 1 to 9 (ablation) | FMA-Net++ · REDS4-ME-5:5 (x4 VSRDB) | PSNR: 28.52 -> 29.24 | Improved |
| add RAFT flow supervision (lambda2 = 1e-4) (ablation) | FMA-Net++ · REDS4-ME-5:5 (x4 VSRDB) | PSNR: 29.07 -> 29.24 | 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.
- FMA-Net* — FMA-Net: Flow-Guided Dynamic Filtering and Iterative Feature Refinement with Multi-Attention for Joint Video Super-Resolution and Deblurring (CVPR 2024) · better on 8 of 8 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
- FMA-Net: Flow-Guided Dynamic Filtering and Iterative Feature Refinement with Multi-Attention for Joint Video Super-Resolution and Deblurring (CVPR 2024 · similarity 0.93)
- Joint Video Multi-Frame Interpolation and Deblurring Under Unknown Exposure Time (CVPR 2023 · similarity 0.90)
- Exploiting Blurry Representations for Event-guided Video Super-Resolution (AAAI 2026 · similarity 0.89)
- Event-Enhanced Blurry Video Super-Resolution (AAAI 2025 · similarity 0.87)
- One-Step Diffusion for Detail-Rich and Temporally Consistent Video Super-Resolution (NeurIPS 2025 · similarity 0.86)
- Blur-aware Spatio-temporal Sparse Transformer for Video Deblurring (CVPR 2024 · 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.
