Research map / 3D vision / Face and shape
JR2Net: Joint Monocular 3D Face Reconstruction and Reenactment
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 |
|---|---|---|---|
| use dense pixel-level conditions and 3D reconstruction guidance for reenactment | JR2Net DCNet · VoxCeleb2 cross-identity | EFL: 2.88 -> 2.50 | Helped |
| use dense pixel-level conditions and 3D reconstruction guidance for reenactment | JR2Net DCNet · VoxCeleb2 cross-identity | FID: 41.48 -> 37.45 | Helped |
| train the expression module using reenactment-generated image-expression pairs | RecAugNet · FaceWarehouse | RMSE: 1.90 -> 1.78 | Helped |
| train RecNet semi-supervised with image-expression pairs and expression loss (ablation) | RecNet · FaceWarehouse | RMSE: 1.90 -> 1.98 | Hurt |
| replace the expression tracker with an empty ETNet (ablation) | RecAugNet · FaceWarehouse | RMSE: 1.89 -> 1.78 | Hurt |
| add generated image-expression pairs to RecNet training data (ablation) | RecNet · FaceWarehouse | RMSE: 1.90 -> 1.92 | Hurt |
| use rendered images instead of dense pixel-level conditions (ablation) | DCNet · VoxCeleb2 same-identity | L1: 0.042 -> 0.044 | Hurt |
| remove illumination control (ablation) | DCNet · VoxCeleb2 same-identity | L1: 0.042 -> 0.048 | Hurt |
| include warped source images as conditional input (ablation) | DCNet · VoxCeleb2 same-identity | L1: 0.045 -> 0.042 | Helped |
Closest papers by meaning
- Face2Face: Real-Time Face Capture and Reenactment of RGB Videos (CVPR 2016 · similarity 0.89)
- FSRT: Facial Scene Representation Transformer for Face Reenactment from Factorized Appearance Head-pose and Facial Expression Features (CVPR 2024 · similarity 0.88)
- VOODOO 3D: Volumetric Portrait Disentanglement For One-Shot 3D Head Reenactment (CVPR 2024 · 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 0.85)
- Learning Dense Correspondence for NeRF-Based Face Reenactment (AAAI 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.