Research map / Video / Video segmentation and tracking
Video Super-Resolution With Temporal Group Attention
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 frame-rate-aware temporal groups with the reference frame in every group (ablation) | TGA · Vid4 and Vimeo-90K-T | PSNR: 27.54 -> 27.59; 37.51 -> 37.59 | Helped |
| Use hierarchical temporal grouping instead of direct temporal stacking (ablation) | TGA · Vid4 and Vimeo-90K-T | PSNR: 27.18 -> 27.59; 37.06 -> 37.59 | Helped |
| Include the reference frame in each temporal group (ablation) | TGA · Vid4 and Vimeo-90K-T | PSNR: 27.47 -> 27.54; 37.46 -> 37.51 | Helped |
| Add group attention (ablation) | TGA · Vid4 and Vimeo-90K-T | PSNR: 27.51 -> 27.59; 37.43 -> 37.59 | Helped |
| Use 7 rather than 5 input frames (ablation) | TGA · Vid4 and Vimeo-90K-T | PSNR: 27.39 -> 27.59; 37.34 -> 37.59 | Helped |
| Add fast spatial alignment (ablation) | TGA · Vimeo-90K-T | PSNR: 37.32 -> 37.59 | Helped |
| Use PyFlow pre-alignment instead of no pre-alignment (ablation) | TGA · Vimeo-90K-T | PSNR: 37.32 -> 35.14 | Hurt |
| Use TGA instead of PFNL | VSR · Vid4, Average (Y) | PSNR: 27.40 -> 27.59 | Helped |
| Use TGA instead of EDVR-L | VSR · Vimeo-90K-T, Y | PSNR: 37.61 -> 37.59 | Hurt |
| Use TGA instead of EDVR-L | VSR · Vimeo-90K-T, RGB | SSIM: 0.9374 -> 0.9387 | Helped |
Closest papers by meaning
- Memory-Augmented Non-Local Attention for Video Super-Resolution (CVPR 2022 · similarity 0.91)
- Real-Time Video Super-Resolution With Spatio-Temporal Networks and Motion Compensation (CVPR 2017 · similarity 0.89)
- VideoGigaGAN: Towards Detail-rich Video Super-Resolution (CVPR 2025 · similarity 0.86)
- EvSTVSR: Event Guided Space-Time Video Super-Resolution (AAAI 2025 · similarity 0.86)
- Enhancing Video Super-Resolution via Implicit Resampling-based Alignment (CVPR 2024 · similarity 0.86)
- Sequence Matters: Harnessing Video Models in 3D Super-Resolution (AAAI 2025 · similarity 0.85)
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.