AMASS
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Human Pose Forecasting | AMASS BMLrub (test) | MPJPE (mm)10.3 | 40 | |
| Human Motion Prediction | AMASS BMLrub | MPJPE (80ms)10.2 | 16 | |
| Motion Prediction | AMASS | MPJPE (80ms)11 | 15 | |
| IK Solver | AMASS (test) | FID0.681 | 12 | |
| Motion Prediction | AMASS | ADE0.3 | 12 | |
| Pose Denoising | AMASS (test) | Delta Q + m2m0.17 | 10 | |
| Stochastic Human Motion Prediction | AMASS (test) | APD0 | 10 | |
| MAV Navigation | AMASS 1 Human scenario (simulation) | Collision Avoidance Rate100 | 9 | |
| Long-term Human Motion Prediction | AMASS (test) | MAE6.2 | 9 | |
| Human Motion Prediction | AMASS (cross-dataset) | ADE0.48 | 9 | |
| MAV Navigation | AMASS 2 Humans scenario (simulation) | Collision Avoidance Rate100 | 8 | |
| Human Motion Reconstruction | AMASS (Protocol 1) | MPJRE2.25 | 8 | |
| Human Pose Forecasting | AMASS (test) | ADE0.311 | 8 | |
| Body estimation | AMASS (test) | MPJPE (mm)119.7 | 8 | |
| Full-body avatar reconstruction | AMASS (S1) | MPJRE2.1 | 8 | |
| Full-body Avatar Reconstruction | AMASS Setting S1 (test) | MPJRE2.53 | 8 | |
| Pose Generation | AMASS (test) | FID0.048 | 8 | |
| 3D Human Pose Reconstruction | AMASS (Setting S2) | MPJPE4.93 | 7 | |
| Human Motion Prediction | AMASS (test) | MPJPE (80ms)9.8 | 7 | |
| Motion Tracking | AMASS (test) | Empirical Joint Position Error (Empjpe)0.2427 | 6 | |
| Text-to-humanoid motion generation | AMASS (test) | FID0 | 6 | |
| Human Pose Prior Evaluation | AMASS (test) | Mean Recall (mm)3.44 | 6 | |
| 3D Body Pose Prediction | AMASS-BMLrub | Accuracy @ 80ms20.6 | 6 | |
| Human Motion Prediction | AMASS 10,000 samples (test) | MAE (L2) @ 100ms0.42 | 6 | |
| Human Motion Forecasting | AMASS (test) | MPJPE (100ms)19 | 6 |
Showing 25 of 62 rows