Cross-Reenactment on VOODOO-XP (test)
0.035MEt3RX-NeMo
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| X-NeMoDimensionality=2D, Training Dataset=Original2025.12 | 0.035 | 0.72 | 0.76 | 0.805 | 0.032 | 0.03 | |
| Portrait4D-v2Dimensionality=3D, Training Dataset=Original2025.12 | 0.035 | 0.79 | 0.589 | 0.886 | 0.029 | 14.5 | |
| GAGAvatarDimensionality=3D, Training Dataset=Original, Inference Optimization=Excluding morphable model fitting2025.12 | 0.034 | 0.77 | 0.654 | 0.888 | 0.025 | 13.67 | |
| Portrait4D-v2†Dimensionality=3D, Training Dataset=Synthetic2025.12 | 0.034 | 0.76 | 0.448 | 0.972 | 0.026 | 14.5 | |
| LivePortraitDimensionality=2D, Training Dataset=Original2025.12 | 0.033 | 0.74 | 0.716 | 0.81 | 0.026 | 78.12 | |
| HYPortraitDimensionality=2D, Training Dataset=Original2025.12 | 0.032 | 0.74 | 0.752 | 0.9 | 0.077 | 0.01 | |
| VOODOO-XPDimensionality=3D, Training Dataset=Original2025.12 | 0.032 | 0.77 | 0.699 | 0.903 | 0.028 | 5.45 | |
| GAGAvatar†Dimensionality=3D, Training Dataset=Synthetic, Inference Optimization=Excluding morphable model fitting2025.12 | 0.03 | 0.67 | 0.578 | 0.944 | 0.038 | 13.67 | |
| InvertAvatarDimensionality=3D, Training Dataset=Original, Inference Optimization=Excluding morphable model fitting2025.12 | 0.028 | 0.8 | 0.565 | 0.891 | 0.049 | 24.3 | |
| Instant Expressive Gaussian Head AvatarDimensionality=3D, Training Dataset=Original2025.12 | 0.028 | 0.75 | 0.771 | 0.745 | 0.028 | 107.31 |