Research map / Video / Video segmentation and tracking
First Order Motion Model for Image Animation
Aliaksandr Siarohin, Stéphane Lathuilière, Sergey Tulyakov, Elisa Ricci, Nicu Sebe
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 first-order local affine motion, occlusion-aware generation, and multiresolution reconstruction instead of prior model-free methods | First Order Motion Model · Tai-Chi-HD | L1: 0.077 -> 0.063 | Improved |
| Use the full method instead of prior model-free methods | First Order Motion Model · VoxCeleb | L1: 0.049 -> 0.043 | Improved |
| Use the full method instead of prior model-free methods | First Order Motion Model · Nemo | L1: 0.018 -> 0.016 | Improved |
| Use the full method instead of prior model-free methods | First Order Motion Model · BAIR | L1: 0.034 -> 0.027 | Improved |
| Add multiresolution pyramid reconstruction loss (ablation) | Baseline · Ablation study | Reconstruction metrics | Mixed |
| Add occlusion-aware generator (ablation) | Pyr. · Ablation study | Reconstruction metrics | Improved |
| Add equivariance constraints on Jacobians (ablation) | Jac. w/o equivariance · Ablation study | Motion estimation quality | Improved |
| Add local affine transformations and the full model components (ablation) | Baseline · Ablation study | MKR: factor of 2.75 | Improved |
Closest papers by meaning
- Motion Representations for Articulated Animation (CVPR 2021 · similarity 0.89)
- Continuous Piecewise-Affine Based Motion Model for Image Animation (AAAI 2024 · similarity 0.87)
- MotionPro: A Precise Motion Controller for Image-to-Video Generation (CVPR 2025 · similarity 0.86)
- Animate Anyone: Consistent and Controllable Image-to-Video Synthesis for Character Animation (CVPR 2024 · similarity 0.86)
- VideoJAM: Joint Appearance-Motion Representations for Enhanced Motion Generation in Video Models (ICML 2025 · similarity 0.86)
- Single Motion Diffusion (ICLR 2024 · 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.
