Research map / Video / Video generation
FFP-300K: Scaling First-Frame Propagation for Generalizable Video Editing
Xijie Huang, Chengming Xu, Donghao Luo, Xiaobin Hu, Peng Tang, Xu Peng, Jiangning Zhang, Chengjie Wang, Yanwei Fu
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 AST-RoPE (ablation) | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench ablation (81-frame model) | VLM score: 7.210 -> 7.542 | Improved |
| add AST-RoPE (ablation) | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench ablation (81-frame model) | PickScore: 20.010 -> 20.354 | Improved |
| add AST-RoPE (ablation) | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench ablation (81-frame model) | DINO temporal consistency: 0.984 -> 0.988 | Improved |
| add self-distillation with identity propagation (on top of AST-RoPE) (ablation) | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench ablation (81-frame model) | VLM score: 7.542 -> 7.600 | Improved |
| add self-distillation with identity propagation (on top of AST-RoPE) (ablation) | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench ablation (81-frame model) | PickScore: 20.354 -> 20.405 | Improved |
| add self-distillation with identity propagation (on top of AST-RoPE) (ablation) | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench ablation (81-frame model) | DINO temporal consistency: 0.988 -> 0.991 | Improved |
| train on FFP-300K with 81 frames instead of 33 frames | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench (125-video FFP subset) | Video text alignment: 25.398 -> 25.925 | Improved |
| train on FFP-300K with 81 frames instead of 33 frames | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench (125-video FFP subset) | VLM score: 7.631 -> 7.600 | No clear effect |
| fine-tune Fun-Control on FFP-300K vs prior FFP methods | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench (125-video FFP subset) | VLM score: 7.341 -> 7.600 | Improved |
| use FFP approach with edited first frame vs Aleph | FreeProp (Fun-Control fine-tuned on FFP-300K) · EditVerseBench (125-video FFP subset) | PickScore: 20.291 -> 20.405 | 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.
- EditVerse — EditVerse: Unifying Image and Video Editing and Generation with In-Context Learning (ICLR 2026) · better on 6 of 6 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
- VINCIE-NExT: Unlocking Video Editing from Images via In-Context Modeling (NeurIPS 2026 · similarity 0.89)
- FADE: Frequency-Aware Diffusion Model Factorization for Video Editing (CVPR 2025 · similarity 0.87)
- Señorita-2M: A High-Quality Instruction-based Dataset for General Video Editing by Video Specialists (NeurIPS 2025 · similarity 0.87)
- Dynamic Motion Blending for Versatile Motion Editing (CVPR 2025 · similarity 0.87)
- Transforming Image Editors into Video Editors (NeurIPS 2026 · similarity 0.87)
- VideoSPatS: Video SPatiotemporal Splines for Disentangled Occlusion, Appearance and Motion Modeling and Editing (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.
