Research map / Video / Video segmentation and tracking
AnimeSR: Learning Real-World Super-Resolution Models for Animation Videos
Yanze Wu, Xintao Wang, Gen Li, Ying Shan
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 |
|---|---|---|---|
| train on AVC-Train instead of ATD-12K (ablation) | AnimeSR · AVC-RealLQ | MANIQA: 0.3309 -> 0.3763 | Improved |
| add one learnable basic operator vs BasicOP-Only (ablation) | AnimeSR · AVC-RealLQ | MANIQA: 0.3554 -> 0.3763 | Improved |
| use tiny LBO instead of large NN-Only degradation (DSGAN-SR) (ablation) | AnimeSR · AVC-RealLQ | MANIQA: 0.3034 -> 0.3763 | Improved |
| NN-Only degradation vs BasicOP-Only (ablation) | AnimeSR · AVC-RealLQ | MANIQA: 0.3554 -> 0.3034 | Hurt |
| multi-scale instead of single-scale recurrent block (ablation) | AnimeSR · AVC-RealLQ | MANIQA: 0.3283 -> 0.3763 | Improved |
| LBO pool of three instead of single LBO-1 (ablation) | AnimeSR · AVC-RealLQ | MANIQA: 0.3763 -> 0.3832 | Improved |
| 3 LBO instead of 1 LBO (ablation) | AnimeSR · AVC-RealLQ | NIQE: 8.1444 -> 8.7088 | Hurt |
| 3 LBO instead of 1 LBO (ablation) | AnimeSR · AVC-RealLQ | MANIQA: 0.3763 -> 0.3832 | Improved |
| AnimeSR vs finetuned RealBasicVSR | AnimeSR · AVC-RealLQ | runtime (ms): 190 -> 32 | Improved |
| AnimeSR vs finetuned BSRGAN | AnimeSR · AVC-RealLQ | runtime (ms): 439 -> 32 | Improved |
Later papers that compared with this method
1 later accepted paper reports AnimeSR as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Learning Data-Driven Vector-Quantized Degradation Model for Animation Video Super-Resolution (ICCV 2023 · 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
- Learning Data-Driven Vector-Quantized Degradation Model for Animation Video Super-Resolution (ICCV 2023 · similarity 0.93)
- Investigating Tradeoffs in Real-World Video Super-Resolution (CVPR 2022 · similarity 0.89)
- APISR: Anime Production Inspired Real-World Anime Super-Resolution (CVPR 2024 · similarity 0.88)
- NegVSR: Augmenting Negatives for Generalized Noise Modeling in Real-world Video Super-Resolution (AAAI 2024 · similarity 0.87)
- VideoGigaGAN: Towards Detail-rich Video Super-Resolution (CVPR 2025 · similarity 0.87)
- NSARM: Next-Scale Autoregressive Modeling for Robust Real-World Image Super-Resolution (NeurIPS 2026 · 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.
