Monocular 3D Human Pose Estimation On Human3
評価指標
Average MPJPE (mm)
PA-MPJPE
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
比較表
| モデル名 | Average MPJPE (mm) | PA-MPJPE |
|---|---|---|
| poseaug-a-differentiable-pose-augmentation | 50.2 | 39.1 |
| 3d-human-pose-estimation-using-spatio-1 | 40.1 | 30.7 |
| ktpformer-kinematics-and-trajectory-prior | 40.1 | - |
| exploiting-temporal-context-for-3d-human-pose | - | - |
| unite-the-people-closing-the-loop-between-3d | - | 80.7 |
| unite-the-people-closing-the-loop-between-3d | - | - |
| srnet-improving-generalization-in-3d-human | 49.9 | - |
| convolutional-mesh-regression-for-single | 74.7 | - |
| diffusion-based-3d-human-pose-estimation-with | 39.5 | - |
| predicting-camera-viewpoint-improves-cross | 52.0 | - |
| neural-body-fitting-unifying-deep-learning | - | - |
| ray3d-ray-based-3d-human-pose-estimation-for | - | - |
| sparseness-meets-deepness-3d-human-pose | 113.01 | - |
| neural-body-fitting-unifying-deep-learning | - | 59.9 |
| poseaug-a-differentiable-pose-augmentation | - | - |
| 3d-human-pose-estimation-in-video-with | 46.8 | - |
| semantic-graph-convolutional-networks-for-3d | 57.6 | - |
| ordinal-depth-supervision-for-3d-human-pose | - | - |
| anatomy-aware-3d-human-pose-estimation-in | 44.1 | - |
| a-dual-source-approach-for-3d-human-pose | 97.39 | - |
| hemlets-pose-learning-part-centric-heatmap-1 | 39.9 | 27.9 |
| 3d-human-pose-estimation-in-the-wild-by | - | - |
| trajectory-space-factorization-for-deep-video | 46.6 | - |
| towards-3d-human-pose-estimation-in-the-wild | 64.9 | - |
| exploiting-temporal-context-for-3d-human-pose | 63.3 | - |
| lifting-from-the-deep-convolutional-3d-pose | 88.39 | - |
| genhmr-generative-human-mesh-recovery | 41.2 | 29.8 |
| camera-distance-aware-top-down-approach-for | - | - |
| 3d-human-pose-estimation-with-spatial-and | 44.3 | - |
| hemlets-pose-learning-part-centric-heatmap-1 | - | - |
| learning-3d-human-pose-from-structure-and | 52.1 | - |
| motionbert-unified-pretraining-for-human | 39.2 | - |
| xnect-real-time-multi-person-3d-human-pose | 63.6 | - |
| motionagformer-enhancing-3d-human-pose | 38.4 | - |
| a-simple-yet-effective-baseline-for-3d-human | 62.9 | - |
| repnet-weakly-supervised-training-of-an | 89.9 | - |
| motionagformer-enhancing-3d-human-pose | 42.5 | - |
| improving-robustness-and-accuracy-via | 44.3 | - |
| generating-multiple-hypotheses-for-3d-human | 52.7 | - |
| end-to-end-recovery-of-human-shape-and-pose | - | - |
| mixste-seq2seq-mixed-spatio-temporal-encoder | 39.8 | - |
| vibe-video-inference-for-human-body-pose-and | 65.6 | - |
| sampling-is-matter-point-guided-3d-human-mesh-1 | 48.3 | 32.9 |
| attention-mechanism-exploits-temporal | 45.1 | - |
| p-stmo-pre-trained-spatial-temporal-many-to | 42.1 | - |
| cascaded-deep-monocular-3d-human-pose-1 | 50.9 | - |
| motionagformer-enhancing-3d-human-pose | 45.1 | - |
| motionbert-unified-pretraining-for-human | 37.5 | - |
| monocular-total-capture-posing-face-body-and | 58.3 | - |
| motionagformer-enhancing-3d-human-pose | 38.4 | - |
| monocular-3d-human-pose-estimation-by-1 | 58.0 | - |