3D Human Keypoint Estimation on Waymo Open Dataset (WOD) 1% (train)
17.2MPJPE (cm)GC-KPL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GC-KPLBackbone=Pre-trained on 200,000 WOD-train, Stage I supervised=true2023.06 | 17.2 | -2.37 | |
| GC-KPLBackbone=Pre-trained on 200,000 WOD-train, Stage I supervised=false2023.06 | 17.8 | -1.77 | |
| GC-KPLBackbone=Pre-trained on 5,000 WOD-train, Stage I supervised=false2023.06 | 17.87 | -1.7 | |
| GC-KPLBackbone=Pre-trained on synthetic only, Stage I supervised=false2023.06 | 18.52 | -1.05 | |
| HUM3DILBackbone=Randomly initialized, Stage I supervised=false2023.06 | 19.57 | — |