3D Human Pose Estimation via Explicit Compositional Depth Maps
n this work, we tackle the problem of estimating 3D human pose in camera space from a monocular image. First, we propose to use densely-generated limb depth maps to ease the learning of body joints depth, which are well aligned with image cues. Then, we design a lifting module from 2D pixel coordinates to 3D camera coordinates which explicitly takes the depth values as inputs, and is aligned with camera perspective projection model. We show our method achieves superior performance on large-scale 3D pose datasets Human3.6M and MPI-INF-3DHP, and sets the new state-of-the-art.
PDF
Datasets
Human3.6M
MPI-INF-3DHP
Results from the Paper
Ranked #21 on
3D Human Pose Estimation
on MPI-INF-3DHP
(PCK metric)
| Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Uses Extra Training Data |
Benchmark |
|---|---|---|---|---|---|---|---|
| 3D Human Pose Estimation | Human3.6M | Explicit Compositional Depth Maps (HRNet-W32 MPII) | Average MPJPE (mm) | 43.2 | # 89 | ||
| Using 2D ground-truth joints | No | # 2 | |||||
| Multi-View or Monocular | Monocular | # 1 | |||||
| PA-MPJPE | 34.6 | # 24 | |||||
| 3D Human Pose Estimation | Human3.6M | Explicit Compositional Depth Maps (ResNet-50 MPII) | Average MPJPE (mm) | 47.3 | # 132 | ||
| Using 2D ground-truth joints | No | # 2 | |||||
| Multi-View or Monocular | Monocular | # 1 | |||||
| PA-MPJPE | 37.3 | # 35 | |||||
| 3D Human Pose Estimation | MPI-INF-3DHP | Explicit Compositional Depth Maps | AUC | 62.4 | # 23 | ||
| PCK | 93.2 | # 21 |
