Research map / Video / Video segmentation and tracking
ART-VSR: Adaptive Rectified Trajectories for One-Step Video Super-Resolution
JianHui Zhang, Chen Fang, Wanghao Wanghao, chaoyu feng, LEI LEI, Jue Wang, Shuaicheng Liu
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 Latent Trajectory Rectifier (ablation) | ART-VSR · REDS | MUSIQ: 46.437 -> 64.965 | Improved |
| add Latent Trajectory Rectifier (ablation) | ART-VSR · REDS | LPIPS: 0.264 -> 0.197 | Improved |
| add Latent Trajectory Rectifier (ablation) | ART-VSR · YouHQ40 | MUSIQ: 42.548 -> 70.917 | Improved |
| add Latent Trajectory Rectifier (ablation) | ART-VSR · VideoLQ | MUSIQ: 23.249 -> 55.272 | Improved |
| add Geometric Projection Loss to LTR (ablation) | ART-VSR · REDS | NIQE: 3.972 -> 2.806 | Improved |
| add Geometric Projection Loss to LTR (ablation) | ART-VSR · REDS | LPIPS: 0.212 -> 0.197 | Improved |
| add Geometric Projection Loss to LTR (ablation) | ART-VSR · MVSR4x | MUSIQ: 48.227 -> 66.563 | Improved |
| use adaptive token-wise timestep map instead of global T_L=0.4 (ablation) | ART-VSR · REDS | LPIPS: 0.231 -> 0.197 | Improved |
| use adaptive token-wise timestep map instead of global T_L=0.4 (ablation) | ART-VSR · REDS | MUSIQ: 42.595 -> 64.965 | Improved |
| add neural branch to heuristic ATE (ablation) | ART-VSR · UDM10 | LPIPS: 0.268 -> 0.242 | Improved |
| add heuristic branch to neural ATE (ablation) | ART-VSR · UDM10 | MUSIQ: 43.987 -> 62.741 | Improved |
| add heuristic branch to neural ATE (ablation) | ART-VSR · MVSR4x | MUSIQ: 45.555 -> 66.563 | Improved |
Closest papers by meaning
- Learning Trajectory-Aware Transformer for Video Super-Resolution (CVPR 2022 · similarity 0.90)
- DTI: Dynamic Trajectory Initialization for Generative Face Video Super-Resolution (ECCV 2026 · similarity 0.90)
- One-Step Diffusion for Detail-Rich and Temporally Consistent Video Super-Resolution (NeurIPS 2025 · similarity 0.88)
- Sequence Matters: Harnessing Video Models in 3D Super-Resolution (AAAI 2025 · similarity 0.87)
- VideoGigaGAN: Towards Detail-rich Video Super-Resolution (CVPR 2025 · similarity 0.87)
- SinSR: Diffusion-Based Image Super-Resolution in a Single Step (CVPR 2024 · 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.
