Head Pose Estimation On Biwi
Metrics
MAE (trained with BIWI data)
MAE (trained with other data)
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | MAE (trained with BIWI data) | MAE (trained with other data) |
|---|---|---|
| a-data-driven-approach-to-improve-3d-head | 2.80 | 4.52 |
| towards-robust-and-unconstrained-full-range | - | 3.39 |
| fine-grained-head-pose-estimation-without | - | 4.89 |
| an-effective-deep-network-for-head-pose | 3.43 | - |
| 6d-rotation-representation-for-unconstrained | 2.66 | 3.47 |
| deep-ordinal-regression-with-label-diversity | 2.54 | - |
| hybrid-coarse-fine-classification-for-head | 3.0174 | - |
| whenet-real-time-fine-grained-estimation-for | - | 3.81 |
| multi-task-head-pose-estimation-in-the-wild-1 | - | 3.66 |
| 2d-image-head-pose-estimation-via-latent | - | 3.519 |
| dad-3dheads-a-large-scale-dense-accurate-and | 3.98 | - |
| face-alignment-across-large-poses-a-3d | - | 19.068 |
| kepler-keypoint-and-pose-estimation-of | - | 13.852 |
| 6dof-head-pose-estimation-through-explicit | - | 2.75 |
| fsa-net-learning-fine-grained-structure | - | 4.00 |
| img2pose-face-alignment-and-detection-via | - | 3.786 |
| whenet-real-time-fine-grained-estimation-for | - | 3.48 |
| on-the-representation-and-methodology-for-1 | - | 4.13 |
| how-far-are-we-from-solving-the-2d-3d-face | - | 7.882 |
| deep-head-pose-estimation-using-synthetic | - | 4.13 |
| towards-robust-and-unconstrained-full-range | - | - |
| single-image-3d-face-reconstruction-under | - | 2.95 |
| dynamic-facial-analysis-from-bayesian | 1.413 | - |
| on-the-power-of-data-augmentation-for-head | - | 3.57 |
| 6dof-head-pose-estimation-through-explicit | - | 2.89 |
| on-the-representation-and-methodology-for-1 | - | 3.98 |
| an-intuitive-and-unconstrained-2d-cube | 4.31 | - |
| fine-grained-head-pose-estimation-without | 4.895 | - |
| img2pose-face-alignment-and-detection-via | - | 4.578 |