3D Human Pose Estimation on Human3.6M (S9, S11)
15.3Average Error (MPJPE Avg)MTF-Transformer
Evaluation results
| Method | Links | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MTF-TransformerMethod Type=Multi-view, Camera Parameters=Without, 2D Detector=GT, T=272021.10 | 15.3 | 15.5 | 17.1 | 13.7 | 15.5 | 14 | 16.2 | 15.8 | 16.5 | 15.8 | 16.1 | 14.5 | 14.5 | 16.9 | 14.3 | 13.7 | — | |
| FinePOSEN=243, Detector=GT2024.05 | 16.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 12.7 | |
| MotionBERTN=243, Detector=GT2024.05 | 16.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| D3DPN=243, Detector=GT2024.05 | 18.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DiffposeN=243, Detector=GT2024.05 | 18.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Zhang et al.Method Type=Multi-view, Camera Parameters=With, 2D Detector=*, T=1, Citation=[23]2021.10 | 19.5 | 17.8 | 19.5 | 17.6 | 20.7 | 19.3 | 16.8 | 18.9 | 20.2 | 25.7 | 20.1 | 19.2 | 20.5 | 17.2 | 20.5 | 17.3 | — | |
| Iskakov et al.Method Type=Multi-view, Camera Parameters=With, 2D Detector=*, T=12021.10 | 20.8 | 19.9 | 20 | 18.9 | 18.5 | 20.5 | 19.4 | 18.4 | 22.1 | 22.5 | 28.7 | 21.2 | 20.8 | 19.7 | 22.1 | 20.2 | — | |
| GLA-GCNN=243, Detector=GT2024.05 | 21 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 17.6 | |
| MixSTEN=243, Detector=GT2024.05 | 21.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Zhang et al.Method Type=Multi-view, Camera Parameters=With, 2D Detector=*, T=1, Citation=[63]2021.10 | 21.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ActionPromptN=243, Detector=GT2024.05 | 22.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FLEXMethod Type=Multi-view, Camera Parameters=Without, 2D Detector=GT, T=272021.10 | 22.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MTF-Transformer+Method Type=Multi-view, Camera Parameters=With, 2D Detector=CPN, T=272021.10 | 25.8 | 23.4 | 25.2 | 23.1 | 24.4 | 27.4 | 28.5 | 22.8 | 25.2 | 28.7 | 36.2 | 25.9 | 23.6 | 26.6 | 22.6 | 22.7 | — | |
| Qiu et al.Method Type=Multi-view, Camera Parameters=With, 2D Detector=*, T=12021.10 | 26.2 | 24 | 26.7 | 23.2 | 24.3 | 24.8 | 22.8 | 24.1 | 28.6 | 32.1 | 26.9 | 31 | 25.6 | 25 | 28 | 24.4 | — | |
| MTF-TransformerMethod Type=Multi-view, Camera Parameters=Without, 2D Detector=CPN, T=272021.10 | 26.2 | 23.1 | 25.4 | 24.7 | 24.5 | 27.9 | 28.3 | 23.9 | 24.6 | 30.7 | 35.7 | 25.8 | 24.2 | 28.4 | 22.8 | 23.1 | — | |
| MTF-Transformer+Method Type=Multi-view, Camera Parameters=With, 2D Detector=CPN, T=12021.10 | 26.5 | 23.8 | 26 | 23.9 | 25 | 28.2 | 29.7 | 23.6 | 25.5 | 30.1 | 37.3 | 26.6 | 24.5 | 27.4 | 23.1 | 23.4 | — | |
| MTF-TransformerMethod Type=Multi-view, Camera Parameters=Without, 2D Detector=CPN, T=27, Special=no added view2021.10 | 26.6 | 24.6 | 25.4 | 24.8 | 24.6 | 28.7 | 29.1 | 23.9 | 25.6 | 31.4 | 36.2 | 26.6 | 24.7 | 28.9 | 23.7 | 23.6 | — | |
| He et al.(IMU)Method Type=Multi-view, Camera Parameters=With, 2D Detector=*, T=12021.10 | 26.9 | 25.7 | 27.7 | 23.7 | 24.8 | 26.9 | 31.4 | 24.9 | 26.5 | 28.8 | 31.7 | 28.2 | 26.4 | 23.6 | 28.3 | 23.5 | — | |
| MTF-TransformerMethod Type=Multi-view, Camera Parameters=Without, 2D Detector=CPN, T=12021.10 | 27.5 | 24.2 | 26.4 | 26.1 | 25.6 | 29.4 | 29.7 | 25.1 | 25.4 | 32.4 | 37.4 | 27.1 | 25.4 | 29.5 | 23.8 | 24.4 | — | |
| P-STMON=243, Detector=GT2024.05 | 29.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Remeli et al.Method Type=Multi-view, Camera Parameters=With, 2D Detector=*, T=12021.10 | 30.2 | 27.3 | 32.1 | 25 | 26.5 | 29.3 | 35.4 | 28.8 | 31.6 | 36.4 | 31.7 | 31.2 | 29.9 | 26.9 | 33.7 | 30.4 | — | |
| MHFormerN=351, Detector=GT2024.05 | 30.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FLEXMethod Type=Multi-view, Camera Parameters=Without, 2D Detector=[22], T=272021.10 | 30.9 | 23.1 | 28.8 | 26.8 | 28.1 | 31.6 | 37.1 | 25.7 | 31.4 | 36.5 | 39.6 | 35 | 29.5 | 35.6 | 26.8 | 26.4 | — | |
| FLEXMethod Type=Multi-view, Camera Parameters=Without, 2D Detector=CPN, T=272021.10 | 31.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FinePOSEN=243, Detector=CPN2024.05 | 31.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 25 | |
| AnatomyN=243, Detector=GT2024.05 | 32.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| D3DPN=243, Detector=CPN2024.05 | 35.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 28.7 | |
| PoseFormerV2N=243, Detector=GT2024.05 | 35.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DiffposeN=243, Detector=CPN2024.05 | 36.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 28.7 | |
| MotionBERTN=243, Detector=SH2024.05 | 37.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Huang et al.Method Type=Multi-view, Camera Parameters=Without, 2D Detector=*, T=12021.10 | 37.5 | 26.8 | 32 | 25.6 | 52.1 | 33.3 | 42.3 | 25.8 | 25.9 | 40.5 | 76.6 | 39.1 | 54.5 | 35.9 | 25.1 | 24.2 | — | |
| TCNN=243, Detector=GT2024.05 | 37.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MixSTEN=243, Detector=HRNet2024.05 | 39.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 30.6 | |
| GazeDH=20, A=ORC2026.01 | 41.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ActionPromptN=243, Detector=CPN2024.05 | 41.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 29.5 | |
| P-STMON=243, Detector=CPN2024.05 | 42.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 34.4 | |
| Cheng et al.Method Type=Monocular, 2D Detector=CPN, T=1282021.10 | 42.9 | 38.3 | 41.3 | 46.1 | 40.1 | 41.6 | 51.9 | 41.8 | 40.9 | 51.5 | 58.4 | 42.2 | 44.6 | 41.7 | 33.7 | 30.1 | — | |
| MHFormerN=351, Detector=CPN2024.05 | 43 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 34.4 | |
| Zhao [71]2026.01 | 43.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AnatomyN=243, Detector=CPN2024.05 | 44.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 35 | |
| DAF-DGDG=true2024.03 | 44.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 34.6 | |
| GLA-GCNN=243, Detector=CPN2024.05 | 44.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 34.8 | |
| Wang et al.Method Type=Monocular, 2D Detector=CPN, T=962021.10 | 44.5 | 40.2 | 42.5 | 42.6 | 41.1 | 46.7 | 56.7 | 41.4 | 42.3 | 56.2 | 60.4 | 46.3 | 42.2 | 46.2 | 31.7 | 31 | — | |
| Zeng et al.Method Type=Monocular, 2D Detector=CPN, T=2432021.10 | 44.8 | 46.6 | 47.1 | 43.9 | 41.6 | 45.8 | 49.6 | 46.5 | 40 | 53.4 | 61.1 | 46.1 | 42.6 | 43.1 | 31.5 | 32.6 | — | |
| Ci et al. (S=10)Method category=Learning Methods, Number of hypotheses (S)=10, 2D detector=Stacked Hourglass2023.07 | 45.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 30.5 | |
| Liu et al.Method Type=Monocular, 2D Detector=CPN, T=2432021.10 | 45.1 | 41.8 | 44.8 | 41.1 | 44.9 | 47.4 | 54.1 | 43.4 | 42.2 | 56.2 | 63.6 | 45.3 | 43.5 | 45.3 | 31.3 | 32.2 | — | |
| PoseFormerV2N=243, Detector=CPN2024.05 | 45.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 35.6 | |
| TCNN=243, Detector=CPN2024.05 | 46.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 36.5 | |
| PavlloSupervision=Supervised2021.09 | 46.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Pavllo et al.Method Type=Monocular, 2D Detector=CPN, T=2432021.10 | 46.8 | 45.2 | 46.7 | 43.3 | 45.6 | 48.1 | 55.1 | 44.6 | 44.3 | 57.3 | 65.8 | 47.1 | 44 | 49 | 32.8 | 33.9 | — | |
| Li et al.Method category=Learning Methods, 2D detector=Stacked Hourglass2023.07 | 47.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 32.7 | |
| CEE-NetVenue=AAAI'23, DG=true2024.03 | 47.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 36.8 | |
| HDVRSupervision=Semi-Sup., Extra Data=No2021.09 | 47.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Diffupose2026.01 | 49.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Gong et al. (f=1)Method category=Learning Methods, Number of frames (f)=1, 2D detector=Stacked Hourglass2023.07 | 49.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 31.6 | |
| Diffpose2026.01 | 49.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GazeDH=20, A=AVG2026.01 | 49.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DH-AUGVenue=ECCV'22, DG=true2024.03 | 49.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 38.3 | |
| Gong et al.Method category=Learning Methods, 2D detector=Stacked Hourglass2023.07 | 50.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 39.1 | |
| PoseAugVenue=TPAMI'23, DG=true2024.03 | 50.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 39.1 | |
| EvoSkeletonVenue=CVPR'20, DG=true2024.03 | 50.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 38 | |
| Ci et al. (S=1)Method category=Learning Methods, Number of hypotheses (S)=1, 2D detector=Stacked Hourglass2023.07 | 51 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ZeDO (S=50)Method category=Optimization Methods, Number of hypotheses (S)=50, 2D detector=Stacked Hourglass2023.07 | 51.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 42.1 | |
| PavlloSupervision=Semi-Sup., Extra Data=No, Ground Truth 2D Poses=true2021.09 | 51.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Zhao [72]2026.01 | 51.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Xu2026.01 | 51.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Liu2026.01 | 52.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Pavllo et al. (f=1)Method category=Learning Methods, Number of frames (f)=1, 2D detector=Stacked Hourglass2023.07 | 52.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 40.9 | |
| VPose (1-frame)Venue=CVPR'19, DG=false2024.03 | 52.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 40.9 | |
| Chen et al.Method Type=Monocular, 2D Detector=CPN, T=12021.10 | 52.7 | 43.8 | 48.6 | 49.1 | 49.8 | 57.6 | 61.5 | 45.9 | 48.3 | 62 | 73.4 | 54.8 | 50.6 | 56 | 43.4 | 45.5 | — | |
| Pavlakos et al.Method Type=Multi-view, Camera Parameters=With, 2D Detector=*, T=12021.10 | 56.9 | 41.2 | 49.2 | 42.8 | 43.4 | 55.6 | 46.9 | 40.3 | 63.7 | 97.6 | 119 | 52.1 | 42.7 | 51.9 | 41.8 | 39.4 | — | |
| ZeDO (S=10)Method category=Optimization Methods, Number of hypotheses (S)=10, 2D detector=Stacked Hourglass2023.07 | 57.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 45.1 | |
| Zhao et al.Method category=Learning Methods, 2D detector=Stacked Hourglass2023.07 | 57.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Hossain2026.01 | 58.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Yang2026.01 | 58.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Sun2026.01 | 59.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| KocabasSupervision=Self-Sup., Extra Data=Multiview2021.09 | 60.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Zhao [70]2026.01 | 60.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Martinez et al.Method category=Learning Methods, 2D detector=Stacked Hourglass2023.07 | 62.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 47.7 | |
| Martinez2026.01 | 62.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MehtaSupervision=Supervised2021.09 | 63.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ZhouSupervision=Weakly-Sup., Extra Data=Yes2021.09 | 64.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ZeDO (S=1)Method category=Optimization Methods, Number of hypotheses (S)=1, 2D detector=Stacked Hourglass2023.07 | 65.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 49 | |
| KunduSupervision=Unsupervised, Extra Data=Yes2021.09 | 67.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ChenSupervision=Unsupervised, Extra Data=Yes2021.09 | 68 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ZanfirSupervision=Supervised2021.09 | 69 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Gu et al.Method category=Optimization Methods, 2D detector=Stacked Hourglass2023.07 | 77.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Li et al.Method category=Optimization Methods, 2D detector=Stacked Hourglass2023.07 | 78.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| HDVRSupervision=Unsupervised, Extra Data=No2021.09 | 82.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Bogo et al.Method category=Optimization Methods, 2D detector=Stacked Hourglass2023.07 | 82.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MartinezSupervision=Supervised2021.09 | 87.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Wang et al.Method category=Optimization Methods, 2D detector=Stacked Hourglass2023.07 | 88 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| WandtSupervision=Supervised2021.09 | 89.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RhodinSupervision=Unsupervised, Extra Data=Multiview2021.09 | 98.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Song et al.Method category=Optimization Methods, 2D detector=Stacked Hourglass2023.07 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 56.4 |