Research map / Video / Video generation
YOSE: You Only Select Essential Tokens for Efficient DiT-based Video Object Removal
Chenyang Wu, Lina Lei, Fan Li, Chunle Guo, Dehong Kong, Xinran Qin, Zhixin Wang, Mingming Cheng, Chongyi Li
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 Lat_mask term to DiffSim (ablation) | YOSE (MiniMax) · DAVIS | PSNR: 27.6164 -> 29.5855 | Improved |
| add Lat_Nis term to DiffSim (ablation) | YOSE (MiniMax) · DAVIS | PSNR: 27.6686 -> 29.5855 | Improved |
| add boundary fusion strategy (ablation) | YOSE (MiniMax) · DAVIS | PSNR: 28.4222 -> 29.5855 | Improved |
| add Lat_mask to DiffSim (ablation) | YOSE (MiniMax) · DAVIS | Dynamic Degree: 0.5444 -> 0.5778 | Improved |
| add fusion strategy (ablation) | YOSE (MiniMax) · DAVIS | Dynamic Degree: 0.5667 -> 0.5778 | Improved |
| add Lat_Nis to DiffSim (ablation) | YOSE (MiniMax) · DAVIS | Aesthetic Quality: 0.4382 -> 0.4432 | Improved |
| add BVI for variable-length multi-batch training (ablation) | YOSE · training time | hours: 11 hours -> about 4 hours | Improved |
| apply YOSE to VACE | VACE · YouTube-VOS | background PSNR: 23.7225 -> 29.1917 | Improved |
| apply YOSE to VACE | VACE · DAVIS | background PSNR: 25.0868 -> 28.2804 | Improved |
| apply YOSE to MiniMax Remover | MiniMax Remover · YouTube-VOS | background PSNR: 30.3296 -> 31.0117 | Improved |
| apply YOSE to MiniMax Remover | MiniMax Remover · DAVIS | background PSNR: 29.3671 -> 29.5855 | Improved |
| apply YOSE to MiniMax Remover | MiniMax Remover · YouTube-VOS | LPIPS: 0.0615 -> 0.0642 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 3 on most shared results.
- MiniMax Remover — MiniMax-Remover: Taming Bad Noise Helps Video Object Removal (NeurIPS 2025) · better on 4 of 6 shared results
- ROSE — ROSE: Remove Objects with Side Effects in Videos (NeurIPS 2025) · better on 4 of 4 shared results
- Propainter — ProPainter: Improving Propagation and Transformer for Video Inpainting (ICCV 2023) · better on 0 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
- AsymRnR: Video Diffusion Transformers Acceleration with Asymmetric Reduction and Restoration (ICML 2025 · similarity 0.88)
- SparseDiT: Token Sparsification for Efficient Diffusion Transformer (NeurIPS 2025 · similarity 0.88)
- MiniMax-Remover: Taming Bad Noise Helps Video Object Removal (NeurIPS 2025 · similarity 0.88)
- Mask^2DiT: Dual Mask-based Diffusion Transformer for Multi-Scene Long Video Generation (CVPR 2025 · similarity 0.87)
- Q-VDiT: Towards Accurate Quantization and Distillation of Video-Generation Diffusion Transformers (ICML 2025 · similarity 0.87)
- Layer- and Timestep-Adaptive Differentiable Token Compression Ratios for Efficient Diffusion Transformers (CVPR 2025 · 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.
