Research map / 3D vision / Face and shape
AU-Aware 3D Face Reconstruction through Personalized AU-Specific Blendshape Learning
Chenyi Kuang, Zijun Cui, Jeffrey O. Kephart, Qiang Ji
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 deformation gradient loss L_G (ablation) | backbone · BU3DFE | mean error (mm): 1.62 -> 1.37 | Improved |
| add deformation gradient loss L_G (ablation) | backbone+L_id · NoW | mean error (mm): 1.84 -> 1.47 | Improved |
| add identity consistency loss L_id (ablation) | backbone · BU3DFE | mean error (mm): 1.62 -> 1.58 | Improved |
| add identity consistency loss L_id (ablation) | backbone · NoW | mean error (mm): SD 1.45 -> 1.56; mean 1.90 -> 1.84 | Mixed |
| add AU label loss L_au-label (ablation) | baseline · BU3DFE | mean error (mm): 1.32 -> 1.18 | Improved |
| add AU label loss L_au-label (ablation) | baseline · NoW | mean error (mm): 1.47 -> 1.39 | Improved |
| add AU relationship prior loss L_au-corr (ablation) | baseline+AU label · BU3DFE | mean error (mm): 1.18 -> 1.11 | Improved |
| add AU relationship prior loss L_au-corr (ablation) | baseline+AU label · NoW | mean error (mm): 1.39 -> 1.33 | Improved |
| add AU label loss L_au-label (ablation) | baseline · BP4D | average F1: 56.0 -> 60.0 | Improved |
| add AU relationship prior L_au-corr (ablation) | baseline+AU label · BP4D | average F1: 60.0 -> 60.8 | Improved |
| add deformation gradient loss L_G (ablation) | backbone · BP4D | average F1: 30.3 -> 51.0 | Improved |
| add identity consistency loss L_id (ablation) | backbone · BP4D | average F1: 30.3 -> 30.4 | No clear effect |
Closest papers by meaning
- Personalized Face Modeling for Improved Face Reconstruction and Motion Retargeting (ECCV 2020 · similarity 0.90)
- AU-Blendshape for Fine-grained Stylized 3D Facial Expression Manipulation (ICCV 2025 · similarity 0.90)
- Learning Person-Specific Animatable Face Models from In-the-Wild Images via a Shared Base Model (CVPR 2025 · similarity 0.86)
- 3D Facial Expressions through Analysis-by-Neural-Synthesis (CVPR 2024 · similarity 0.86)
- TEASER: Token Enhanced Spatial Modeling for Expressions Reconstruction (ICLR 2025 · similarity 0.86)
- OFER: Occluded Face Expression Reconstruction (CVPR 2025 · 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.
