3D Hand Pose Estimation On Freihand
Metriken
PA-F@15mm
PA-F@5mm
PA-MPJPE
PA-MPVPE
Ergebnisse
Leistungsergebnisse verschiedener Modelle zu diesem Benchmark
Vergleichstabelle
| Modellname | PA-F@15mm | PA-F@5mm | PA-MPJPE | PA-MPVPE |
|---|---|---|---|---|
| weakly-supervised-mesh-convolutional-hand | 0.966 | 0.614 | 8.4 | 8.6 |
| sr-affine-high-quality-3d-hand-model | 0.977 | 0.707 | 6.7 | 6.9 |
| Modell 3 | 0.981 | 0.724 | 6.3 | 6.5 |
| deformable-mesh-transformer-for-3d-human-mesh | 0.984 | 0.743 | 6.2 | 6.4 |
| fastvit-a-fast-hybrid-vision-transformer | 0.981 | 0.722 | 6.6 | 6.7 |
| freihand-a-dataset-for-markerless-capture-of | 0.934 | 0.516 | 11.0 | 10.9 |
| end-to-end-human-pose-and-mesh-reconstruction | 0.984 | 0.731 | 6.5 | 6.3 |
| monocular-expressive-body-regression-through | 0.918 | 0.484 | 12.2 | 11.8 |
| hamba-single-view-3d-hand-reconstruction-with | 0.992 | 0.806 | 5.7 | 5.3 |
| wilor-end-to-end-3d-hand-localization-and | 0.993 | 0.825 | 5.5 | 5.1 |
| mmhmr-generative-masked-modeling-for-hand | 0.991 | 0.801 | 5.5 | 5.4 |
| hand-image-understanding-via-deep-multi-task | 0.974 | 0.699 | 7.1 | 7.3 |
| freihand-a-dataset-for-markerless-capture-of | 0.935 | 0.529 | - | 10.7 |
| mobrecon-mobile-friendly-hand-mesh | 0.986 | 0.784 | 5.7 | 5.8 |
| towards-accurate-alignment-in-real-time-3d | 0.981 | 0.724 | 6.7 | 6.7 |
| self-supervised-3d-hand-pose-estimation-from | - | - | 6.6 | - |
| monocular-real-time-full-body-capture-with | - | - | 15.7 | - |
| pose2pose-3d-positional-pose-guided-3d | 0.971 | 0.664 | 7.7 | 7.7 |
| pose2mesh-graph-convolutional-network-for-3d | 0.969 | 0.674 | 7.7 | 7.8 |
| a-probabilistic-attention-model-with | 0.987 | 0.767 | 6.2 | 6.1 |
| cross-attention-of-disentangled-modalities | 0.983 | 0.687 | 6.5 | 7.1 |
| learning-joint-reconstruction-of-hands-and | 0.908 | 0.436 | - | 13.2 |
| reconstructing-hands-in-3d-with-transformers | 0.990 | 0.785 | 6.0 | 5.7 |
| collaborative-regression-of-expressive-bodies | 0.919 | 0.468 | 12 | 12.1 |
| i2l-meshnet-image-to-lixel-prediction-network-1 | 0.973 | 0.681 | 7.4 | 7.6 |
| camera-space-hand-mesh-recovery-via-semantic | 0.977 | 0.715 | 6.9 | 7.0 |
| handos-3d-hand-reconstruction-in-one-stage | 0.991 | 0.812 | 5.0 | 5.3 |
| mesh-graphormer | 0.986 | 0.764 | 5.9 | 6.0 |
| hhmr-holistic-hand-mesh-recovery-by-enhancing | - | - | 5.8 | 5.8 |
| a-simple-baseline-for-efficient-hand-mesh | 0.986 | 0.772 | 5.7 | 6.0 |
| 3d-hand-shape-and-pose-from-images-in-the | 0.898 | 0.435 | - | 13.0 |
| sampling-is-matter-point-guided-3d-human-mesh-1 | 0.984 | 0.720 | 6.1 | 6.6 |