Research map / Video / Video generation
Unified In-Context Video Editing
Zixuan Ye, Xuanhua He, Quande Liu, Qiulin Wang, Xintao Wang, Pengfei Wan, Di ZHANG, Kun Gai, Qifeng Chen, Wenhan Luo
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 task-aware RoPE and condition bias (D4 vs D1) (ablation) | UNIC · Re-Camera Control | TransErr: 8.972 -> 5.667 | Improved |
| add condition bias only to sequential-RoPE model (ablation) | UNIC · Re-Camera Control | RotErr: 2.501 -> 1.428 | Improved |
| add task-aware RoPE only (ablation) | UNIC · Re-Camera Control | CamMC: 13.119 -> 6.498 | Improved |
| add condition bias on top of task-aware RoPE (ablation) | UNIC · Stylization | ArtFID: 32.768 -> 37.619 | Hurt |
| add condition bias on top of task-aware RoPE (ablation) | UNIC · ID Swap | CLIP-I: 0.710 -> 0.725 | Improved |
| staged training starting with camera instead of joint training (ablation) | UNIC · Re-Camera Control | RotErr: 2.287 -> 1.275 | Improved |
| staged training starting with camera instead of joint training (ablation) | UNIC · Re-Camera Control | TransErr: 9.694 -> 5.667 | Improved |
| easy-to-hard instead of hard-to-easy order (ablation) | UNIC · Re-Camera Control | RotErr: 1.275 -> 1.398 | Hurt |
| joint training instead of staged (ablation) | UNIC · Stylization | CSD-Score: 0.259 -> 0.298 | Mixed |
| unify tasks instead of task-specific camera model (ablation) | UNIC · Re-Camera Control | RotErr: 1.472 -> 1.275 | Improved |
| unify tasks instead of task-specific style model (ablation) | UNIC · Stylization | CSD-Score: 0.234 -> 0.259 | Improved |
| unify tasks instead of task-specific style model (ablation) | UNIC · Stylization | CFSD: 0.096 -> 0.107 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- StyleMaster — StyleMaster: Stylize Your Video with Artistic Generation and Translation (CVPR 2025) · better on 1 of 2 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
- VicEdit: Learning to Edit Videos from Visual In-Context Examples (NeurIPS 2026 · similarity 0.89)
- VINCIE-NExT: Unlocking Video Editing from Images via In-Context Modeling (NeurIPS 2026 · similarity 0.88)
- Señorita-2M: A High-Quality Instruction-based Dataset for General Video Editing by Video Specialists (NeurIPS 2025 · similarity 0.87)
- VideoDirector: Precise Video Editing via Text-to-Video Models (CVPR 2025 · similarity 0.87)
- VEU-Bench: Towards Comprehensive Understanding of Video Editing (CVPR 2025 · similarity 0.87)
- VE-Bench: Subjective-Aligned Benchmark Suite for Text-Driven Video Editing Quality Assessment (AAAI 2025 · 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.
