3D Human Pose Estimation On Human36M
评估指标
Average MPJPE (mm)
PA-MPJPE
评测结果
各个模型在此基准测试上的表现结果
比较表格
| 模型名称 | Average MPJPE (mm) | PA-MPJPE |
|---|---|---|
| a-lightweight-graph-transformer-network-for | 64.3 | 45.4 |
| srnet-improving-generalization-in-3d-human | 44.8 | - |
| mhformer-multi-hypothesis-transformer-for-3d | 43 | - |
| mocap-guided-data-augmentation-for-3d-pose | - | 87.3 |
| probabilistic-monocular-3d-human-pose | 61.8 | - |
| modulated-graph-convolutional-network-for-3d | 49.4 | - |
| kama-3d-keypoint-aware-body-mesh-articulation | - | 40.2 |
| gla-gcn-global-local-adaptive-graph | 21.0 | 17.6 |
| exploiting-temporal-information-for-3d-pose | 58.5 | - |
| weakly-supervised-generative-network-for | 81.1 | - |
| regular-splitting-graph-network-for-3d-human | 47 | 38.6 |
| learning-to-estimate-3d-human-pose-and-shape | - | 75.9 |
| hemlets-pose-learning-part-centric-heatmap-1 | 45.1 | - |
| optimizing-network-structure-for-3d-human | 52.7 | - |
| 3d-human-pose-machines-with-self-supervised | 63.67 | - |
| lightweight-multi-view-3d-pose-estimation | 21.0 | - |
| real-time-multi-view-3d-human-pose-estimation | 23.5 | - |
| remocap-disentangled-representation-learning | 52.8 | 33.4 |
| sparseness-meets-deepness-3d-human-pose | 113.01 | - |
| unihpe-towards-unified-human-pose-estimation | 50.5 | 36.2 |
| lifting-from-the-deep-convolutional-3d-pose | 88.39 | - |
| 模型 22 | 15.6 | - |
| dual-networks-based-3d-multi-person-pose | 34.95 | - |
| adapted-human-pose-monocular-3d-human-pose | - | 85.1 |
| epipolar-transformers | 19.0 | - |
| mug-multi-human-graph-network-for-3d-mesh | 61.9 | 48.5 |
| end-to-end-human-pose-and-mesh-reconstruction | 54 | 36.7 |
| lifting-transformer-for-3d-human-pose | 45.4 | - |
| a-unified-deep-framework-for-joint-3d-pose | 42.4 | - |
| human-pose-estimation-in-space-and-time-using | 119 | - |
| learnable-human-mesh-triangulation-for-3d | 30.56 | - |
| convformer-parameter-reduction-in-transformer | 29.8 | - |
| learning-3d-human-dynamics-from-video | 83.7 | 56.9 |
| denserac-joint-3d-pose-and-shape-estimation | 76.8 | 48 |
| mixste-seq2seq-mixed-spatio-temporal-encoder | 25.9 | - |
| 3d-human-mesh-estimation-from-virtual-markers-1 | 47.3 | 32 |
| anatomy-aware-3d-human-pose-estimation-in | 32.3 | - |
| ddt-a-diffusion-driven-transformer-based | 73.1 | 48.6 |
| a-dual-source-approach-for-3d-pose-estimation | 97.39 | 108.3 |
| motionagformer-enhancing-3d-human-pose | 19.4 | - |
| improving-robustness-and-accuracy-via | 30.1 | - |
| back-to-the-future-joint-aware-temporal-deep | 45.9 | - |
| human-pose-regression-with-residual-log | 36.3 | - |
| towards-3d-human-pose-estimation-in-the-wild | 64.9 | - |
| hstformer-hierarchical-spatial-temporal | 42.7 | 33.7 |
| motionagformer-enhancing-3d-human-pose | 28.1 | - |
| exploiting-spatial-temporal-relationships-for | 50.6 | - |
| 3d-human-pose-and-shape-regression-with | 57.7 | 40.5 |
| 3d-human-pose-estimation-with-2d-marginal | 57 | 40.4 |
| posern-a-2d-pose-refinement-network-for-bias | 38.4 | - |
| chirality-nets-for-human-pose-regression | 46.7 | - |
| capturing-the-motion-of-every-joint-3d-human | 54.9 | 38.4 |
| harvesting-multiple-views-for-marker-less-3d | 56.9 | - |
| not-all-tokens-are-equal-human-centric-visual | 62.9 | 42.8 |
| spgnet-spatial-projection-guided-3d-human | 45.3 | - |
| ppt-token-pruned-pose-transformer-for | 24.4 | - |
| vibe-video-inference-for-human-body-pose-and | 65.6 | 41.4 |
| lifting-monocular-events-to-3d-human-poses | 116.4 | - |
| differentiable-dynamics-for-articulated-3d | 33.4 | 21.9 |
| xnect-real-time-multi-person-3d-human-pose | 63.6 | - |
| convformer-parameter-reduction-in-transformer | 43.2 | - |
| adaptively-multi-view-and-temporal-fusing | 49.4 | - |
| finepose-fine-grained-prompt-driven-3d-human | 16.7 | - |
| i2l-meshnet-image-to-lixel-prediction-network-1 | 55.7 | 41.7 |
| ponet-robust-3d-human-pose-estimation-via | 56.1 | 39.8 |
| unsupervised-3d-pose-estimation-with | 68.0 | - |
| human-mesh-recovery-from-monocular-images-via | 59.1 | 42.4 |
| consensus-based-optimization-for-3d-human | 52 | - |
| hybrik-a-hybrid-analytical-neural-inverse | 54.4 | 33.6 |
| localization-with-sampling-argmax | 49.5 | 39.1 |
| enhanced-3d-human-pose-estimation-from-videos | 44.8 | - |
| posenet3d-unsupervised-3d-human-shape-and | 47.0 | - |
| a-real-time-human-pose-measurement-system-for | 18.2 | - |
| srnet-improving-generalization-in-3d-human | 49.9 | - |
| metrabs-metric-scale-truncation-robust | 49.3±0.7 | - |
| learning-local-recurrent-models-for-human | 61.9 | 42.5 |
| humans-in-4d-reconstructing-and-tracking | 44.8 | 33.6 |
| attention-mechanism-exploits-temporal | 34.7 | - |
| poseaug-a-differentiable-pose-augmentation | 50.8 | - |
| higher-order-implicit-fairing-networks-for-3d | 54.8 | - |
| multi-task-deep-learning-for-real-time-3d | 48.6 | - |
| conditional-directed-graph-convolution-for-3d | 41.1 | - |
| double-chain-constraints-for-3d-human-pose | 32.4 | - |
| total-capture-3d-human-pose-estimation-fusing | 57.0 | - |
| motion-projection-consistency-based-3d-human | 35.4 | - |
| teaching-independent-parts-separately-tips | 43.5 | - |
| learning-skeletal-graph-neural-networks-for | 47.9 | - |
| xformer-fast-and-accurate-monocular-3d-body | 52.6 | 35.2 |
| back-to-optimization-diffusion-based-zero | 51.4 | 42.1 |
| ktpformer-kinematics-and-trajectory-prior | 33.0 | 26.2 |
| mixste-seq2seq-mixed-spatio-temporal-encoder | 21.6 | - |
| interweaved-graph-and-attention-network-for | 47.7 | - |
| posetriplet-co-evolving-3d-human-pose | 78 | 51.8 |
| chirality-nets-for-human-pose-regression | 51.4 | - |
| metric-scale-truncation-robust-heatmaps-for | 49.3±0.7 | - |
| generating-multiple-hypotheses-for-3d-human | 49.6 | - |
| self-attentive-3d-human-pose-and-shape | 58.9 | 38.7 |
| flex-parameter-free-multi-view-3d-human | 24.8 | - |
| 3d-human-pose-regression-using-graph | 52.8 | - |
| predicting-peoples-3d-poses-from-short-1 | 125.28 | - |
| lcr-net-localization-classification | 87.7 | 71.6 |
| neural-descent-for-visual-3d-human-pose-and | 91.8 | 66 |
| voxelkeypointfusion-generalizable-multi-view | 64.3 | - |
| 3d-human-pose-estimation-via-intuitive | 60.6 | 41.8 |
| dc-gnet-deep-mesh-relation-capturing-graph | 72.3 | 42.4 |
| evopose-a-recursive-transformer-for-3d-human | 42.83 | - |
| ray3d-ray-based-3d-human-pose-estimation-for | 84.4 | - |
| smap-single-shot-multi-person-absolute-3d | 54.1 | - |
| learnable-human-mesh-triangulation-for-3d | 17.59 | - |
| gator-graph-aware-transformer-with-motion | 48.8 | 31.2 |
| learning-to-fuse-2d-and-3d-image-cues-for | 69.73 | - |
| fbi-pose-towards-bridging-the-gap-between-2d | 56.5 | - |
| capturing-humans-in-motion-temporal-attentive | 69.4 | 47.4 |
| embodied-scene-aware-human-pose-estimation | 103.4 | 73.7 |
| motion-projection-consistency-based-3d-human | 32.5 | - |
| multi-initialization-optimization-network-for | 56.88 | 41.59 |
| probabilistic-monocular-3d-human-pose | 44.3 | - |
| thundr-transformer-based-3d-human | 87 | 62.2 |
| upose3d-uncertainty-aware-3d-human-pose | 26.4 | 23.4 |
| motion-projection-consistency-based-3d-human | 47.4 | - |
| 3d-human-pose-estimation-2d-pose-estimation | 82.72 | - |
| diffusion-based-3d-human-pose-estimation-with | 19.6 | - |
| lifting-transformer-for-3d-human-pose | 46.9 | - |
| double-chain-constraints-for-3d-human-pose | 46.1 | - |
| exploiting-spatial-temporal-relationships-for | 49.1 | - |
| occlusion-aware-networks-for-3d-human-pose | 42.9 | - |
| mhformer-multi-hypothesis-transformer-for-3d | 30.5 | - |
| 3d-human-pose-estimation-with-spatial-and | 44.3 | - |
| convolutional-mesh-regression-for-single | 51.9 | 51.9 |
| poseaug-a-differentiable-pose-augmentation | 38.2 | - |
| exploiting-temporal-information-for-3d-human | 51.9 | 42.0 |
| 3d-human-pose-estimation-using-mobius-graph | 52.1 | - |
| ssp-net-scalable-sequential-pyramid-networks | 50.2 | - |
| cliff-carrying-location-information-in-full | 47.1 | - |
| repnet-weakly-supervised-training-of-an | 50.9 | - |
| a-simple-yet-effective-baseline-for-3d-human | 45.5 | - |
| poseformerv2-exploring-frequency-domain-for | 45.2 | - |
| hemlets-pose-learning-part-centric-heatmap-1 | 39.9 | 27.9 |
| 3d-human-pose-estimation-with-siamese | 61.1 | - |
| graph-stacked-hourglass-networks-for-3d-human | 51.9 | - |
| skelformer-markerless-3d-pose-and-shape | 25.2 | 20.6 |
| rethinking-pose-in-3d-multi-stage-refinement | 52.8 | - |
| tessetrack-end-to-end-learnable-multi-person | 44.6 | - |
| camera-distance-aware-top-down-approach-for | 54.4 | - |
| w-hmr-human-mesh-recovery-in-world-space-with | 45.5 | 30.2 |
| spgnet-spatial-projection-guided-3d-human | 33.4 | - |
| learning-dynamical-human-joint-affinity-for | 31.2 | - |
| pliks-a-pseudo-linear-inverse-kinematic | 47 | 34.5 |
| metapose-fast-3d-pose-from-multiple-views | 49 | - |
| skeleton-transformer-networks-3d-human-pose | 69.95 | 61.4 |
| cross-view-fusion-for-3d-human-pose | 31.17 | - |
| how-robust-is-3d-human-pose-estimation-to | 56.1 | - |
| posenet3d-unsupervised-3d-human-shape-and | 59.4 | - |
| thundr-transformer-based-3d-human | 55 | 39.8 |
| 3d-human-pose-estimation-via-explicit | 43.2 | 34.6 |
| consensus-based-optimization-for-3d-human | 45 | - |
| context-modeling-in-3d-human-pose-estimation | 43.4 | - |
| ray3d-ray-based-3d-human-pose-estimation-for | 34.4 | - |
| sampling-is-matter-point-guided-3d-human-mesh-1 | 48.3 | 32.9 |
| beyond-static-features-for-temporally | 73.6 | 52 |
| 3d-human-pose-estimation-via-explicit | 47.3 | 37.3 |
| live-stream-temporally-embedded-3d-human-body | 68.6 | 47.1 |
| exploiting-temporal-context-for-3d-human-pose | 77.8 | 41.6 |
| cross-view-fusion-for-3d-human-pose | 26.21 | - |
| absposelifter-absolute-3d-human-pose-lifting | 38.38 | 28.78 |
| compositional-human-pose-regression | 59.1 | 48.3 |
| adaptively-multi-view-and-temporal-fusing | 50.7 | - |
| motion-projection-consistency-based-3d-human | 44.8 | - |
| ordinal-depth-supervision-for-3d-human-pose | 56.2 | - |
| mixste-seq2seq-mixed-spatio-temporal-encoder | 42.4 | - |
| anatomy-aware-3d-human-pose-estimation-in | 44.1 | - |
| bodynet-volumetric-inference-of-3d-human-body | 51.6 | - |
| body-meshes-as-points | - | 51.3 |
| exploiting-temporal-context-for-3d-human-pose | 63.3 | - |
| absposelifter-absolute-3d-human-pose-lifting | 52.5 | 39.1 |
| hopfir-hop-wise-graphformer-with-intragroup | 48.5 | - |
| solopose-one-shot-kinematic-3d-human-pose | 26.0 | 11.5 |
| gast-net-graph-attention-spatio-temporal | 46.2 | 36 |
| hybrik-x-hybrid-analytical-neural-inverse | 47 | 29.8 |
| 3d-human-pose-estimation-from-a-single-image | - | 76.5 |
| ivt-an-end-to-end-instance-guided-video | 40.2 | - |
| cascaded-deep-monocular-3d-human-pose-1 | 50.9 | - |
| dynamic-graph-reasoning-for-multi-person-3d | 41.3 | 27.3 |
| structured-prediction-of-3d-human-pose-with | 125.0 | - |
| pedrecnet-multi-task-deep-neural-network-for | 52.7 | - |
| p-stmo-pre-trained-spatial-temporal-many-to | 42.1 | 34.4 |
| pose-conditioned-joint-angle-limits-for-3d | - | 181.1 |
| 190505754 | 20.8 | - |
| diffupose-monocular-3d-human-pose-estimation | 50.7 | - |
| 3d-human-pose-estimation-in-video-with | 46.8 | 36.5 |
| keep-it-smpl-automatic-estimation-of-3d-human | - | 82.3 |
| jointformer-single-frame-lifting-transformer | 50.5 | - |
| learning-temporal-3d-human-pose-estimation | 50.6 | - |
| holopose-holistic-3d-human-reconstruction-in | 50.42 | 36.82 |
| 3d-human-pose-estimation-using-convolutional | 117.34 | - |
| motionagformer-enhancing-3d-human-pose | 26.5 | - |
| repnet-weakly-supervised-training-of-an | 89.9 | - |
| encoder-decoder-with-multi-level-attention | 56.4 | 38.7 |
| staf-3d-human-mesh-recovery-from-video-with | 70.4 | 44.5 |
| refined-temporal-pyramidal-compression-and | 40.1 | 32 |
| unsupervised-cross-modal-alignment-for-multi | 67.9 | - |
| deep-two-stream-video-inference-for-human | 60.5 | 39.3 |
| poseaug-a-differentiable-pose-augmentation | 36.9 | - |
| spatio-temporal-tendency-reasoning-for-human | 67.8 | 46.6 |
| monocular-3d-human-pose-estimation-in-the | 72.88 | - |
| motionagformer-enhancing-3d-human-pose | 17.3 | - |
| crossformer-cross-spatio-temporal-transformer | 43.7 | - |
| exploiting-spatial-temporal-relationships-for | 48.8 | - |
| sparseness-meets-deepness-3d-human-pose | - | 106.7 |
| deciwatch-a-simple-baseline-for-10x-efficient | 53.1 | - |
| a-simple-yet-effective-baseline-for-3d-human | 62.9 | - |
| tape-temporal-attention-based-probabilistic | 60 | 39.5 |
| graphmlp-a-graph-mlp-like-architecture-for-3d | 48 | - |
| ktpformer-kinematics-and-trajectory-prior | 18.1 | - |
| mug-multi-human-graph-network-for-3d-mesh | 50.3 | 38.5 |
| learning-3d-human-shape-and-pose-from-dense | 54.6 | 42.9 |
| compressed-volumetric-heatmaps-for-multi | 51.1 | 43.4 |
| orinet-a-fully-convolutional-network-for-3d | 63.7 | 46.6 |
| arts-semi-analytical-regressor-using | 51.6 | 36.6 |
| predicting-camera-viewpoint-improves-cross | 52 | 42.5 |
| 3d-human-pose-estimation-using-spatio-1 | 40.1 | 30.7 |
| 3d-human-mesh-regression-with-dense-1 | - | 39.3 |
| end-to-end-recovery-of-human-shape-and-pose | 87.97 | 58.1 |
| self-supervised-3d-human-pose-estimation-with | 62.0 | - |
| hspace-synthetic-parametric-humans-animated | 53.3 | 39 |
| real-time-multi-view-3d-human-pose-estimation | 29.8 | - |
| vnect-real-time-3d-human-pose-estimation-with | 80.5 | - |
| semantic-estimation-of-3d-body-shape-and-pose | 49.9 | - |
| lifting-transformer-for-3d-human-pose | 28.5 | - |
| poseaug-a-differentiable-pose-augmentation | 50.2 | - |
| gla-gcn-global-local-adaptive-graph | 44.4 | 34.8 |
| exemplar-fine-tuning-for-3d-human-pose | - | 46.8 |
| 2d3d-pose-estimation-and-action-recognition | 53.2 | - |
| ray3d-ray-based-3d-human-pose-estimation-for | 49.7 | - |
| real-time-monocular-full-body-capture-in | 53.9 | 42.4 |
| trajectory-optimization-for-physics-based | 84 | 56 |
| 190505754 | 49.9 | - |
| pymaf-x-towards-well-aligned-full-body-model | 54.2 | 37.2 |
| probabilistic-modeling-for-human-mesh | - | 41.2 |
| hybrik-transformer | 47.5 | 29.5 |
| lightweight-multi-view-3d-pose-estimation | 30.2 | - |
| a-simple-yet-effective-baseline-for-3d-human | 67.5 | - |
| mixste-seq2seq-mixed-spatio-temporal-encoder | 40.9 | - |
| monocular-total-capture-posing-face-body-and | 58.3 | - |
| adaptively-multi-view-and-temporal-fusing | 29.4 | - |
| generating-multiple-hypotheses-for-3d-human | 52.7 | 42.6 |
| graph-and-temporal-convolutional-networks-for | 40.9 | 30.4 |
| neural-body-fitting-unifying-deep-learning | - | 59.9 |
| tessetrack-end-to-end-learnable-multi-person | 18.7 | - |
| self-supervised-learning-of-3d-human-pose | 51.83 | - |
| trajectory-space-factorization-for-deep-video | 46.6 | - |
| geometry-biased-transformer-for-robust-multi | 26.0 | - |
| maximum-margin-structured-learning-with-deep | 121.31 | - |
| monocular-3d-human-pose-estimation-by | - | 79.5 |
| 3d-human-pose-estimation-in-the-wild-by | 58.6 | 37.7 |
| weakly-supervised-3d-human-pose-learning-via | 56.1 | - |
| 3d-human-pose-estimation-with-spatio-temporal | 21.3 | - |
| 3d-human-pose-estimation-using-mobius-graph | 36.2 | - |
| kinematic-aware-hierarchical-attention | 35.4 | - |
| generalizing-monocular-3d-human-pose | 37.6 | - |
| the-best-of-both-worlds-combining-model-based | 54.7 | - |
| hdformer-high-order-directed-transformer-for | 42.6 | - |
| posegu-3d-human-pose-estimation-with-novel | 49.6 | - |
| hierarchical-graph-networks-for-3d-human-pose | 51.8 | - |
| 3d-human-motion-estimation-via-motion | 76 | 53.2 |
| hdformer-high-order-directed-transformer-for | 40.3 | - |
| 3d-human-pose-estimation-in-video-with | 51.8 | 40 |
| dual-networks-based-3d-multi-person-pose | 49.31 | - |
| p-stmo-pre-trained-spatial-temporal-many-to | 29.3 | - |
| srnet-improving-generalization-in-3d-human | 32 | - |
| monocular-3d-human-pose-estimation-by-1 | 46.8 | - |
| potter-pooling-attention-transformer-for | 56.5 | 35.1 |
| transfusion-cross-view-fusion-with | 25.8 | - |
| tcpformer-learning-temporal-correlation-with | 15.5 | - |
| learning-to-regress-bodies-from-images-using-1 | 60.9 | 40.3 |
| deep-kinematic-pose-regression | 107.26 | - |
| global-to-local-modeling-for-video-based-3d | 67 | 46.3 |
| cross-attention-of-disentangled-modalities | 52.2 | 33.7 |
| improving-robustness-and-accuracy-via | 44.3 | - |
| lifting-transformer-for-3d-human-pose | 44 | - |
| learning-to-reconstruct-3d-human-pose-and | - | 41.1 |
| jointformer-single-frame-lifting-transformer | 34 | - |
| joint-3d-human-shape-recovery-from-a-single | 61.2 | 35.4 |
| propagating-lstm-3d-pose-estimation-based-on | 52.8 | - |
| fusionformer-exploiting-the-joint-motion | 48.7 | - |
| towards-generalization-of-3d-human-pose | 61.3 | - |
| motion-guided-3d-pose-estimation-from-videos | 25.6 | - |
| sparse-representation-for-3d-shape-estimation | - | 106.7 |
| temporal-aware-refinement-for-video-based | 45.6 | 33.3 |
| semantic-graph-convolutional-networks-for-3d | 57.6 | - |
| solopose-one-shot-kinematic-3d-human-pose | 38.9 | 29.9 |
| pose2mesh-graph-convolutional-network-for-3d | 64.9 | 48.7 |
| generalizing-monocular-3d-human-pose | 58 | - |
| integral-human-pose-regression | 49.6 | 40.6 |
| heater-an-efficient-and-unified-network-for | 49.9 | 32.8 |
| graph-stacked-hourglass-networks-for-3d-human | 35.8 | - |
| motion-guided-3d-pose-estimation-from-videos | 42.6 | - |
| unite-the-people-closing-the-loop-between-3d | - | 80.7 |
| simpoe-simulated-character-control-for-3d | 56.7 | 41.6 |
| posetriplet-co-evolving-3d-human-pose | 68.2 | - |
| htnet-human-topology-aware-network-for-3d | 47.6 | 38.6 |
| shape-aware-human-pose-and-shape | 44.4 | - |
| flex-parameter-free-multi-view-3d-human | 30.9 | - |
| diffupose-monocular-3d-human-pose-estimation | 49.4 | - |
| skelformer-markerless-3d-pose-and-shape | 33.5 | 27.8 |
| p-stmo-pre-trained-spatial-temporal-many-to | 44.1 | - |
| gast-net-graph-attention-spatio-temporal | 45.7 | 35.9 |
| motionbert-unified-pretraining-for-human | 16.9 | - |
| conditional-directed-graph-convolution-for-3d | 22.7 | - |
| diffpose-toward-more-reliable-3d-pose | 36.9 | - |
| crossformer-cross-spatio-temporal-transformer | 28.3 | - |
| learning-dynamical-human-joint-affinity-for | 45.3 | - |
| unihpe-towards-unified-human-pose-estimation | 53.6 | 40.9 |
| camerapose-weakly-supervised-monocular-3d | 38.87 | - |
| single-shot-multi-person-3d-pose-estimation | 69.9 | - |
| pc-hmr-pose-calibration-for-3d-human-mesh | 47.9 | 37.3 |
| beyond-static-features-for-temporally | 62.3 | 41.1 |
| synthetic-training-for-accurate-3d-human-pose | - | 55.4 |
| deep-monocular-3d-human-pose-estimation-via | 48.7 | - |
| learning-temporal-3d-human-pose-estimation | 43.0 | - |
| enhanced-3d-human-pose-estimation-from-videos | 33.4 | - |
| canonpose-self-supervised-monocular-3d-human | 74.3 | - |
| learning-skeletal-graph-neural-networks-for | 30.4 | - |
| hmor-hierarchical-multi-person-ordinal | 48.6 | 30.5 |
| graphmlp-a-graph-mlp-like-architecture-for-3d | 35.1 | - |
| gast-net-graph-attention-spatio-temporal | 49 | 37.4 |
| lifting-transformer-for-3d-human-pose | 43.7 | - |
| camera-distance-aware-top-down-approach-for | 53.3 | - |
| mpt-mesh-pre-training-with-transformers-for | 45.3 | 31.7 |
| deepfuse-an-imu-aware-network-for-real-time | 37.5 | - |
| epipolar-transformers | 26.9 | - |
| zolly-zoom-focal-length-correctly-for | 49.4 | 32.3 |
| self-supervised-3d-human-pose-estimation-via | 62.4 | - |
| stochastic-modeling-for-learnable-human-pose | 29.1 | - |
| learning-pose-grammar-to-encode-human-body | 60.4 | 45.7 |
| synthetic-occlusion-augmentation-with | 54.2 | - |
| adaptively-multi-view-and-temporal-fusing | 28.5 | - |
| srnet-improving-generalization-in-3d-human | 33.9 | - |
| compositional-human-pose-regression | 92.4 | 67.5 |
| consensus-based-optimization-for-3d-human | 39 | - |
| 190505754 | 17.7 | - |
| spatial-temporal-graph-convolutional-networks-1 | 57.4 | - |
| 3d-human-pose-estimation-with-2d-marginal | 55.4 | 39 |
| mixste-seq2seq-mixed-spatio-temporal-encoder | 39.8 | - |
| camera-distortion-aware-3d-human-pose-1 | 51.4 | - |
| semantic-graph-convolutional-networks-for-3d | 43.8 | - |
| smpler-taming-transformers-for-monocular-3d | 45.2 | 32.4 |
| attention-mechanism-exploits-temporal | 45.1 | - |
| 3d-lfm-lifting-foundation-model | 31.89 | - |
| 3d-human-pose-estimation-with-spatial-and | 31.3 | - |
| mesh-graphormer | 51.2 | 34.5 |
| learning-3d-human-pose-from-structure-and | 52.1 | 36.3 |
| recurrent-3d-pose-sequence-machines | 73.1 | - |
| ellipbody-a-light-weight-and-part-based | - | 47.6 |
| adafuse-adaptive-multiview-fusion-for | 19.5 | - |
| optimizing-network-structure-for-3d-human | 36.3 | - |
| coarse-to-fine-volumetric-prediction-for | 71.9 | 51.9 |