Research map / Video / Video generation
MOFA-Video: Controllable Image Animation via Generative Motion Field Adaptions in Frozen Image-to-Video Diffusion Model
Muyao Niu, Xiaodong Cun, Xintao Wang, Yong Zhang, Ying Shan, Yinqiang Zheng
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 explicit feature warping with motion fields instead of sparse conditioning only (ablation) | MOFA-Adapter · WebVid test, trajectory animation | FVD: 184.27 -> 86.76 | Improved |
| add sparse-to-dense motion generator before warping (ablation) | MOFA-Adapter · WebVid test, trajectory animation | LPIPS: 0.2376 -> 0.2274 | Improved |
| add sparse-to-dense motion generator before warping (ablation) | MOFA-Adapter · WebVid test, trajectory animation | FVD: 81.80 -> 86.76 | Hurt |
| finetune the adapter after flow-based reconstruction pretraining (ablation) | MOFA-Adapter · WebVid test, trajectory animation | FVD: 102.17 -> 86.76 | Improved |
| use periodic sampling with overlapped frame groups for long videos (ablation) | MOFA-Video · Long video generation (qualitative) | temporal consistency | Improved |
| use domain-specific landmark adapter instead of the trajectory adapter for portraits (ablation) | MOFA-Adapter · Portrait animation (qualitative) | facial naturalness | Improved |
| MOFA-Video vs DragNUWA for trajectory animation | MOFA-Video · WebVid test | FVD: 91.38 -> 86.76 | Improved |
Closest papers by meaning
- Video Diffusion Models are Training-free Motion Interpreter and Controller (NeurIPS 2024 · similarity 0.88)
- Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance (NeurIPS 2025 · similarity 0.88)
- Learn the Force We Can: Enabling Sparse Motion Control in Multi-Object Video Generation (AAAI 2024 · similarity 0.88)
- Motion Prompting: Controlling Video Generation with Motion Trajectories (CVPR 2025 · similarity 0.88)
- MotionClone: Training-Free Motion Cloning for Controllable Video Generation (ICLR 2025 · similarity 0.87)
- Consistent and Controllable Image Animation with Motion Diffusion Models (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.
