3D Human Pose Estimation On Mpi Inf 3Dhp
Metriken
MPJPE
PCK
Ergebnisse
Leistungsergebnisse verschiedener Modelle zu diesem Benchmark
Vergleichstabelle
| Modellname | MPJPE | PCK |
|---|---|---|
| canonpose-self-supervised-monocular-3d-human | 104 | 77 |
| 3d-human-pose-estimation-in-the-wild-by | - | 69.0 |
| spatio-temporal-tendency-reasoning-for-human | 95.3 | - |
| graph-stacked-hourglass-networks-for-3d-human | - | 80.1 |
| deep-two-stream-video-inference-for-human | 93.4 | - |
| mixste-seq2seq-mixed-spatio-temporal-encoder | 54.9 | 94.4 |
| probabilistic-monocular-3d-human-pose | - | 84.3 |
| self-attentive-3d-human-pose-and-shape | 94.3 | 90.1 |
| ponet-robust-3d-human-pose-estimation-via | 115.0 | 76.1 |
| posetriplet-co-evolving-3d-human-pose | 79.5 | 89.1 |
| crossformer-cross-spatio-temporal-transformer | 76.3 | 89.1 |
| weakly-supervised-3d-human-pose-learning-via | 110.8 | 80.2 |
| srnet-improving-generalization-in-3d-human | - | 77.6 |
| repnet-weakly-supervised-training-of-an | 92.5 | 81.8 |
| absposelifter-absolute-3d-human-pose-lifting | - | 84 |
| 3d-human-pose-estimation-with-spatial-and | 77.1 | 88.6 |
| 3d-human-pose-estimation-with-spatio-temporal | 23.1 | 98.7 |
| w-hmr-human-mesh-recovery-in-world-space-with | 83.2 | - |
| fusionformer-exploiting-the-joint-motion | 28.2 | 97.9 |
| finepose-fine-grained-prompt-driven-3d-human | 26.2 | 98.9 |
| vnect-real-time-3d-human-pose-estimation-with | - | 79.4 |
| vnect-real-time-3d-human-pose-estimation-with | 124.7 | 76.6 |
| leveraging-mocap-data-for-human-mesh-recovery | 97.4 | - |
| capturing-humans-in-motion-temporal-attentive | 96.7 | - |
| learning-temporal-3d-human-pose-estimation | 93.0 | 81.0 |
| posegu-3d-human-pose-estimation-with-novel | 79.1 | 86.3 |
| beyond-static-features-for-temporally | 97.4 | - |
| ddt-a-diffusion-driven-transformer-based | 97.8 | - |
| generating-multiple-hypotheses-for-3d-human | - | 67.9 |
| predicting-camera-viewpoint-improves-cross | 90.3 | 84.3 |
| learnable-human-mesh-triangulation-for-3d | 33.7 | 99.37 |
| hybrik-a-hybrid-analytical-neural-inverse | 91.0 | 87.5 |
| motionagformer-enhancing-3d-human-pose | 16.2 | 98.2 |
| beyond-static-features-for-temporally | 97.3 | - |
| 3d-human-motion-estimation-via-motion | 96.4 | - |
| unsupervised-3d-pose-estimation-with | - | 71.1 |
| motion-guided-3d-pose-estimation-from-videos | 68.1 | 86.9 |
| poseaug-a-differentiable-pose-augmentation | 71.1 | 89.2 |
| end-to-end-recovery-of-human-shape-and-pose | 124.2 | 72.9 |
| learnable-human-mesh-triangulation-for-3d | 45.87 | 96.59 |
| consensus-based-optimization-for-3d-human | 112.1 | 80.6 |
| convformer-parameter-reduction-in-transformer | 53.6 | 96.4 |
| motionagformer-enhancing-3d-human-pose | 19.2 | 98.2 |
| weakly-supervised-generative-network-for | - | 79.3 |
| diffusion-based-3d-human-pose-estimation-with | 29.7 | 97.7 |
| staf-3d-human-mesh-recovery-from-video-with | 92.4 | - |
| arts-semi-analytical-regressor-using | 71.8 | - |
| hstformer-hierarchical-spatial-temporal | 28.3 | 98 |
| monocular-3d-human-pose-estimation-in-the | - | 64.7 |
| repnet-weakly-supervised-training-of-an | 97.8 | 82.5 |
| mixste-seq2seq-mixed-spatio-temporal-encoder | 57.9 | 94.2 |
| learning-skeletal-graph-neural-networks-for | - | 82.1 |
| global-to-local-modeling-for-video-based-3d | 93.9 | - |
| 3d-human-pose-estimation-using-spatio-1 | - | 84.1 |
| orinet-a-fully-convolutional-network-for-3d | - | 64.6 |
| ordinal-depth-supervision-for-3d-human-pose | - | 71.9 |
| conditional-directed-graph-convolution-for-3d | 42.5 | 97.9 |
| global-adaptation-meets-local-generalization | 61.3 | 92.1 |
| metrabs-metric-scale-truncation-robust | 74.9±1.4 | 90.6±0.4 |
| tcpformer-learning-temporal-correlation-with | 17.8 | 98.7 |
| encoder-decoder-with-multi-level-attention | 83.6 | - |
| poseformerv2-exploring-frequency-domain-for | 27.8 | 97.9 |
| poseaug-a-differentiable-pose-augmentation | 76.6 | - |
| context-modeling-in-3d-human-pose-estimation | - | 80.5 |
| modulated-graph-convolutional-network-for-3d | - | 86.1 |
| ktpformer-kinematics-and-trajectory-prior | 16.7 | 98.9 |
| dc-gnet-deep-mesh-relation-capturing-graph | 97.2 | - |
| 3d-human-pose-estimation-via-explicit | - | 93.2 |
| 3d-human-shape-and-pose-from-a-single-low | 103.36 | - |
| hybrik-x-hybrid-analytical-neural-inverse | 91 | 87.1 |
| exemplar-fine-tuning-for-3d-human-pose | - | - |
| htnet-human-topology-aware-network-for-3d | - | 86.7 |
| mhformer-multi-hypothesis-transformer-for-3d | 58 | 93.8 |
| tcpformer-learning-temporal-correlation-with | 15 | 99.0 |
| back-to-optimization-diffusion-based-zero | 55.2 | 93 |
| trajectory-space-factorization-for-deep-video | 79.8 | 83.6 |
| learning-to-reconstruct-3d-human-pose-and | - | 92.5 |
| ssp-net-scalable-sequential-pyramid-networks | 96.8 | 83.2 |
| motionagformer-enhancing-3d-human-pose | 18.2 | 98.3 |
| anatomy-aware-3d-human-pose-estimation-in | 79.1 | 87.8 |
| learning-to-regress-bodies-from-images-using-1 | 104.7 | - |
| double-chain-constraints-for-3d-human-pose | - | 87.5 |
| denserac-joint-3d-pose-and-shape-estimation | 114.2 | 76.9 |
| learning-to-reconstruct-3d-human-pose-and | 105.2 | 76.4 |
| xformer-fast-and-accurate-monocular-3d-body | 109.8 | - |
| poseaug-a-differentiable-pose-augmentation | 73 | 88.6 |
| refined-temporal-pyramidal-compression-and | 40.5 | 98.8 |
| single-shot-multi-person-3d-pose-estimation | 122.2 | 75.2 |
| p-stmo-pre-trained-spatial-temporal-many-to | 32.2 | 97.9 |
| live-stream-temporally-embedded-3d-human-body | 96.2 | - |
| ikol-inverse-kinematics-optimization-layer | 88.8 | 87.9 |
| hybrik-transformer | 86.2 | 88.6 |
| monocular-3d-human-pose-estimation-in-the | 117.6 | 75.7 |
| tape-temporal-attention-based-probabilistic | 94 | - |
| hemlets-pose-learning-part-centric-heatmap-1 | - | 75.3 |
| pliks-a-pseudo-linear-inverse-kinematic | 67.6 | - |
| generalizing-monocular-3d-human-pose | - | 71.2 |
| poseaug-a-differentiable-pose-augmentation | 73.2 | - |
| cascaded-deep-monocular-3d-human-pose-1 | 99.7 | 81.2 |
| ray3d-ray-based-3d-human-pose-estimation-for | 46.6 | - |
| gla-gcn-global-local-adaptive-graph | 27.76 | 98.53 |
| learning-dynamical-human-joint-affinity-for | 76 | - |
| diffpose-toward-more-reliable-3d-pose | 29.1 | 98 |
| temporal-aware-refinement-for-video-based | 85.9 | - |
| teaching-independent-parts-separately-tips | - | 78 |
| anatomy-aware-3d-human-pose-estimation-in | 78.8 | 87.9 |
| hierarchical-graph-networks-for-3d-human-pose | - | 85.2 |
| self-supervised-learning-of-3d-human-pose | 108.99 | 77.5 |
| regular-splitting-graph-network-for-3d-human | - | 85.6 |
| out-of-domain-human-mesh-reconstruction-via | 101.5 | 79.5 |
| motionagformer-enhancing-3d-human-pose | 17.1 | 98.3 |
| 3d-human-pose-estimation-with-2d-marginal | 91.3 | 85.4 |
| posenet3d-unsupervised-3d-human-shape-and | 102.4 | 81.9 |
| xnect-real-time-multi-person-3d-human-pose | 98.4 | 82.8 |
| vibe-video-inference-for-human-body-pose-and | 96.6 | 89.3 |
| learning-local-recurrent-models-for-human | 94.6 | - |