Research map / Video / Video segmentation and tracking
TRaM-VSR: Importance-Aware Token Routing and Merging for One-Step Diffusion Video Super-Resolution
Sicheng Gao, Yixuan Liu, Tong Shen, Zhuyun Zhou, Zongwei Wu, Radu Timofte
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 importance-aware routing instead of random routing (ablation) | TRaM-VSR · SPMCS (ablation) | LPIPS: 0.3187 -> 0.2883 | Improved |
| add global stream to single-stream importance routing (ablation) | TRaM-VSR · SPMCS (ablation) | LPIPS: 0.3391 -> 0.2883 | Improved |
| add global stream to single-stream importance routing (ablation) | TRaM-VSR · SPMCS (ablation) | MUSIQ: 63.28 -> 67.27 | Improved |
| add global stream (ablation) | TRaM-VSR · SPMCS (ablation) | PSNR: 23.08 -> 23.18 | No clear effect |
| add global stream (ablation) | TRaM-VSR · SPMCS (ablation) | E*_warp: 0.9839 -> 0.9720 | Mixed |
| fuse temporal cue with text similarity instead of text only (ablation) | TRaM-VSR · SPMCS (ablation) | LPIPS: 0.3450 -> 0.2883 | Improved |
| fuse text similarity with temporal cue instead of temporal cue only (ablation) | TRaM-VSR · SPMCS (ablation) | MUSIQ: 59.49 -> 67.27 | Improved |
| use temporal cue only instead of fusion (ablation) | TRaM-VSR · SPMCS (ablation) | E*_warp: 0.9720 -> 0.9068 | Improved |
| use offline planner for routed intervals instead of heuristic placements (ablation) | TRaM-VSR · speed-quality tradeoff | quality at fixed speedup | Improved |
| add identity restoration at route exit (ablation) | TRaM-VSR · SPMCS | structural robustness | 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.
- SeedVR2-7B — SeedVR: Seeding Infinity in Diffusion Transformer Towards Generic Video Restoration (CVPR 2025) · better on 7 of 10 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
- Motion-Guided Latent Diffusion for Temporally Consistent Real-world Video Super-resolution (ECCV 2024 · similarity 0.89)
- One-Step Diffusion for Detail-Rich and Temporally Consistent Video Super-Resolution (NeurIPS 2025 · similarity 0.89)
- PS-SR: Pseudo-Single-Step Video Super-Resolution via Speculative Diffusion (CVPR 2026 · similarity 0.89)
- Upscale-A-Video: Temporal-Consistent Diffusion Model for Real-World Video Super-Resolution (CVPR 2024 · similarity 0.89)
- PatchVSR: Breaking Video Diffusion Resolution Limits with Patch-wise Video Super-Resolution (CVPR 2025 · similarity 0.88)
- SeeClear: Semantic Distillation Enhances Pixel Condensation for Video Super-Resolution (NeurIPS 2024 · similarity 0.88)
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.
