Research map / Video / Video segmentation and tracking
Animating Arbitrary Objects via Deep Motion Transfer
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 motion-specific keypoints and dense motion transfer instead of X2Face | Video reconstruction · Tai-Chi | L1: 0.068 -> 0.050 | Helped |
| Use motion-specific keypoints and dense motion transfer instead of X2Face | Video reconstruction · Nemo | L1: 0.022 -> 0.017 | Helped |
| Use motion-specific keypoints and dense motion transfer instead of X2Face | Video reconstruction · BAIR | L1: 0.069 -> 0.025 | Helped |
| Use Monkey-Net instead of the strongest listed baseline | Image-to-video translation · Tai-Chi | FID: 54.83 -> 19.75 | Helped |
| Use Monkey-Net instead of the strongest listed baseline | Image-to-video translation · Nemo | FID: 27.27 -> 11.97 | Helped |
| Use Monkey-Net instead of the strongest listed baseline | Image-to-video translation · BAIR | FID: 57.90 -> 23.20 | Helped |
| Add dense optical flow deformation (ablation) | Monkey-Net · Tai-Chi video reconstruction | L1: 0.057 -> 0.050 | Helped |
| Add a residual motion field (ablation) | Monkey-Net · Tai-Chi video reconstruction | L1: 0.051 -> 0.050 | Helped |
| Add part-based coarse optical flow (ablation) | Monkey-Net · Tai-Chi video reconstruction | L1: 0.052 -> 0.050 | Helped |
| Estimate keypoint covariances instead of using fixed variance (ablation) | Monkey-Net · Tai-Chi video reconstruction | L1: 0.054 -> 0.050 | Helped |
| Provide the source image as input to the motion network (ablation) | Monkey-Net · Tai-Chi video reconstruction | L1: 0.051 -> 0.050 | Helped |
Closest papers by meaning
- MotionPro: A Precise Motion Controller for Image-to-Video Generation (CVPR 2025 · similarity 0.87)
- MotionDreamer: One-to-Many Motion Synthesis with Localized Generative Masked Transformer (ICLR 2025 · similarity 0.86)
- DisPose: Disentangling Pose Guidance for Controllable Human Image Animation (ICLR 2025 · similarity 0.85)
- Learn the Force We Can: Enabling Sparse Motion Control in Multi-Object Video Generation (AAAI 2024 · similarity 0.85)
- Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance (NeurIPS 2025 · similarity 0.85)
- Animate Anyone: Consistent and Controllable Image-to-Video Synthesis for Character Animation (CVPR 2024 · similarity 0.85)
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.