Research map / Video / Video segmentation and tracking
A Codec Information Assisted Framework for Efficient Compressed Video Super-Resolution
Hengsheng Zhang, Xueyi Zou, Jiaming Guo, Youliang Yan, Rong Xie, Li Song
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 MV-based alignment | FRVSR · REDS4 (H.264 CRF23) | PSNR: 27.34 -> 27.77 | Improved |
| add MV-based alignment | RLSP · REDS4 (H.264 CRF23) | PSNR: 27.47 -> 27.95 | Improved |
| add MV-based alignment | RSDN · REDS4 (H.264 CRF23) | PSNR: 27.62 -> 28.02 | Improved |
| replace MV alignment with SpyNet optical flow alignment | RSDN+MV · REDS4 (H.264 CRF18) | PSNR: 29.37 -> 29.59 | Improved |
| replace MV alignment with SpyNet optical flow alignment | RSDN+MV · REDS4 (H.264 CRF28) | PSNR: 26.50 -> 26.51 | No clear effect |
| replace original optical flow model of FRVSR with MV alignment | FRVSR · REDS4 (H.264 CRF23) | params and runtime: 2.59M -> 0.84M params, 24 -> 20 ms | Improved |
| align features instead of the HR image with MVs (ablation) | recurrent baseline · REDS4 (H.264 CRF23) | PSNR: 27.83 -> 28.00 | Improved |
| add MV image alignment to unaligned baseline (ablation) | recurrent baseline · REDS4 (H.264 CRF23) | PSNR: 27.56 -> 27.83 | Improved |
| add Residual-informed sparse processing to MV-aligned baseline | Baseline+MV · REDS4 (H.264 CRF23) | PSNR and sparse rate: 28.00 -> 27.72, sparse rate 0.75 | Mixed |
| add Residual-informed sparse processing to MV-aligned baseline | Baseline+MV · REDS4 (H.264 CRF28) | PSNR: 26.47 -> 26.15 | Hurt |
| use Residual mask instead of CNN mask at test time (ablation) | MV+Res · REDS4 (H.264 CRF18) | PSNR: 28.82 -> 29.03 | Improved |
| use Residual mask instead of CNN mask at test time (ablation) | MV+Res · REDS4 (H.264 CRF28) | PSNR: 26.08 -> 26.15 | Improved |
Closest papers by meaning
- Compression-Aware Video Super-Resolution (CVPR 2023 · similarity 0.88)
- Compressed-Domain-Aware Online Video Super-Resolution (CVPR 2026 · similarity 0.88)
- Recurrent Back-Projection Network for Video Super-Resolution (CVPR 2019 · similarity 0.88)
- Frame-Recurrent Video Super-Resolution (CVPR 2018 · similarity 0.88)
- Efficient Video Compression via Content-Adaptive Super-Resolution (ICCV 2021 · similarity 0.88)
- Video Super-Resolution via Bidirectional Recurrent Convolutional Networks (TPAMI 2017 · similarity 0.87)
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.
