Research map / 3D vision / Face and shape
Towards Accurate Facial Motion Retargeting with Identity-Consistent and Expression-Exclusive Constraints
Read the paperCheck 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 |
|---|---|---|---|
| add identity-consistent constraint (ablation) | CPEM · FEAFA test set | MAE: 0.495 -> 0.269 | Helped |
| add expression-exclusive constraint to identity-consistent CPEM (ablation) | CPEM · FEAFA test set | MAE: 0.269 -> 0.244 | Helped |
| add both identity-consistent and expression-exclusive constraints (ablation) | CPEM · FEAFA test set | MAE: 0.495 -> 0.244 | Helped |
| use λidc=1000 rather than λidc=1 (ablation) | CPEM · FEAFA test set | MAE: 0.491 -> 0.244 | Helped |
| increase λidc from 100 to 1000 (ablation) | CPEM · FEAFA test set | MAE: 0.286 -> 0.244 | Helped |
| increase λidc from 1000 to 10000 (ablation) | CPEM · FEAFA test set | MAE: 0.244 -> 0.279 | Hurt |
| set λexp to 1 rather than 10 (ablation) | CPEM · FEAFA test set | MAE: 0.244 -> 0.260 | Hurt |
| increase λexp from 10 to 100 (ablation) | CPEM · FEAFA test set | MAE: 0.244 -> 0.255 | Hurt |
| use CPEM instead of the strongest listed baseline 3DDFA-v2 | AFLW2000-3D · AFLW2000-3D | NME (%) mean: 3.65 vs. 3.64 | No clear effect |
| use CPEM instead of 3DDFA-v2 | MICC Florence · MICC Florence | geometric reconstruction error (mm): 1.94±0.52 vs. 2.08±0.59; 2.10±0.50 vs. 2.09±0.54; 1.98±0.49 vs. 2.02±0.54 | Mixed |
| use CPEM instead of 3DDFA-v2 | FaceWarehouse · FaceWarehouse | geometric reconstruction error (mm): 2.26±0.43 -> 2.03±0.33 | Helped |
| use CPEM rather than MS-SFN, RingNet, or Personalized | facial motion retargeting · FEAFA test set | MAE: 0.496, 0.390, 0.397 -> 0.244 | Helped |
Closest papers by meaning
- 3D Facial Expressions through Analysis-by-Neural-Synthesis (CVPR 2024 · similarity 0.88)
- TEASER: Token Enhanced Spatial Modeling for Expressions Reconstruction (ICLR 2025 · similarity 0.87)
- FaithfulFaces: Pose-Faithful Facial Identity Preservation for Text-to-Video Generation (NeurIPS 2026 · similarity 0.86)
- MagicPose: Realistic Human Poses and Facial Expressions Retargeting with Identity-aware Diffusion (ICML 2024 · similarity 0.86)
- FSRT: Facial Scene Representation Transformer for Face Reenactment from Factorized Appearance Head-pose and Facial Expression Features (CVPR 2024 · similarity 0.86)
- Enhancing Identity-Deformation Disentanglement in StyleGAN for One-Shot Face Video Re-Enactment (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.