3D Human Pose Estimation on Human3.6M Protocol 1 (test)
17.5Dir. Error (Protocol 1)Reddy et al.
Evaluation results
| Method | Links | ||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Reddy et al.Input Type=Multi-view, Extrinsic Camera Parameters=Given, Exploit Temporal Information=true2021.05 | 17.5 | 19.6 | 17.2 | 18.3 | 18.3 | 18 | 18 | 20.5 | 20.3 | — | 17.7 | 17.2 | 19 | 18.9 | — | 18.7 | 17.8 | — | 19.4 | — | |
| RTPCAInput Source=GT, Video-based=true2023.09 | 18.5 | 19.5 | 18.5 | 19.1 | 19.6 | 21.6 | 19.1 | 26.2 | 25.3 | — | 22.3 | 19.3 | 12.5 | 18.5 | — | 19.5 | 12.8 | — | 19.1 | — | |
| Iskakov et al.Input Type=Multi-view, Extrinsic Camera Parameters=Given2021.05 | 19.9 | 20 | 18.9 | 18.5 | 20.5 | 18.4 | 22.1 | 22.5 | 28.7 | — | 19.4 | 20.8 | 22.1 | 19.7 | — | 20.8 | 20.2 | — | 21.2 | — | |
| GLA-GCNinput_source=GT, temporal_information=true, T=243, intermediate_3D_reconstruction=true2023.07 | 20.1 | 21.2 | 20 | 19.6 | 21.5 | 23.3 | 19.8 | 27 | 29.4 | — | 26.7 | 20.1 | 12.8 | 19.2 | — | 21 | 13.8 | — | 20.8 | — | |
| MixSTET (number of input frames)=243, Temporal information usage=true, Transformer-based architecture=true, Input=2D Ground Truth Keypoints2022.03 | 21.6 | 22 | 20.4 | 21 | 20.8 | 24.7 | 21.9 | 26.9 | 24.9 | 21.2 | 24.3 | 21.5 | 14.7 | 20.8 | — | 21.6 | 15.7 | — | — | — | |
| Zhang et al.input_source=GT, temporal_information=true, intermediate_3D_reconstruction=true2023.07 | 21.6 | 22 | 20.4 | 21 | 20.8 | 24.7 | 21.9 | 26.9 | 24.9 | — | 24.3 | 21.5 | 14.7 | 20.8 | — | 21.6 | 15.7 | — | 21.2 | — | |
| Zhang et al.Input Source=GT, Video-based=true2023.09 | 21.6 | 22 | 20.4 | 21 | 20.8 | 24.7 | 21.9 | 26.9 | 24.9 | — | 24.3 | 21.5 | 14.7 | 20.8 | — | 21.6 | 15.7 | — | 21.2 | — | |
| Wang et al.T (number of input frames)=96, Input=2D Ground Truth Keypoints2022.03 | 23 | 25.7 | 22.8 | 22.6 | 24.1 | 24.9 | 24.5 | 31.1 | 35 | 25.6 | 30.6 | 24.3 | 19.8 | 25.1 | — | 25.6 | 18.4 | — | — | — | |
| Wang et al.input_source=GT, temporal_information=true, intermediate_3D_reconstruction=true2023.07 | 23 | 25.7 | 22.8 | 22.6 | 24.1 | 24.9 | 24.5 | 31.1 | 35 | — | 30.6 | 24.3 | 19.8 | 25.1 | — | 25.6 | 18.4 | — | 25.6 | — | |
| Qiu et al.Input Type=Multi-view, Extrinsic Camera Parameters=Given, Extra Training Data=true2021.05 | 24 | 26.7 | 23.2 | 24.3 | 24.8 | 24.1 | 28.6 | 32.1 | 26.9 | — | 22.8 | 25.6 | 28 | 25 | — | 26.2 | 24.4 | — | 31 | — | |
| Ma et al.Input Type=Multi-view, Extrinsic Camera Parameters=Given, Exploit Temporal Information=true2021.05 | 24.4 | 26.4 | 23.4 | 24.1 | 25.2 | 24.7 | 33.8 | 29.8 | 26.4 | — | 23.2 | 24.2 | 26.1 | 23.2 | — | 25.8 | 23.3 | — | 26.8 | — | |
| MixSTET (number of input frames)=81, Temporal information usage=true, Transformer-based architecture=true, Input=2D Ground Truth Keypoints2022.03 | 25.6 | 27.8 | 24.5 | 25.7 | 24.9 | 28.6 | 27.4 | 29.9 | 29 | 26.1 | 29.9 | 25 | 18.7 | 25.2 | — | 25.9 | 19.9 | — | — | — | |
| Pipeline4Bounding box source=Proposed detection and tracking algorithm2026.01 | 25.61 | 33.44 | 35.51 | 31.47 | 63.73 | 25.44 | 29.68 | 137.46 | 111.77 | 46.28 | 28.62 | 26.7 | 27.72 | 30.67 | — | 45.36 | 26.24 | — | — | — | |
| He et al.Input Type=Multi-view, Extrinsic Camera Parameters=Given2021.05 | 25.7 | 27.7 | 23.7 | 24.8 | 26.9 | 24.9 | 26.5 | 28.8 | 31.7 | — | 31.4 | 26.4 | 28.3 | 23.6 | — | 26.9 | 23.5 | — | 28.2 | — | |
| GLA-GCNinput_source=GT, temporal_information=true, T=2432023.07 | 26.5 | 27.2 | 29.2 | 25.4 | 28.2 | 29.5 | 26.9 | 37.8 | 39.9 | — | 31.7 | 27 | 20.5 | 27.3 | — | 28.5 | 20.8 | — | 29.9 | — | |
| Li et al.input_source=GT, temporal_information=true, intermediate_3D_reconstruction=true2023.07 | 27.1 | 29.4 | 26.5 | 27.1 | 28.6 | 30.7 | 26.8 | 38.2 | 34.7 | — | 33 | 29.8 | 19.1 | 26.8 | — | 28.5 | 19.8 | — | 29.1 | — | |
| Iskakov et al.Input Type=Multi-view, Extrinsic Camera Parameters=Not Given, Non-ep-free algorithm=true, Parameter perturbation=3%2021.05 | 27.6 | 30.3 | 29 | 29.4 | 26.9 | 27.4 | 34.8 | 39.1 | 54 | — | 36.5 | 30.7 | 26.2 | 36.2 | — | 33.1 | 28.4 | — | 34.4 | — | |
| Li et al.input_source=GT, temporal_information=true2023.07 | 27.7 | 32.1 | 29.1 | 28.9 | 30 | 33 | 31.2 | 37 | 39.3 | — | 33.9 | 31 | 22.2 | 29.4 | — | 30.5 | 23 | — | 30 | — | |
| Li et al.Input Source=GT, Video-based=true2023.09 | 27.7 | 32.1 | 29.1 | 28.9 | 30 | 33 | 31.2 | 37 | 39.3 | — | 33.9 | 31 | 22.2 | 29.4 | — | 30.5 | 23 | — | 30 | — | |
| Shan et al.input_source=GT, temporal_information=true2023.07 | 28.5 | 30.1 | 28.6 | 27.9 | 29.8 | 31.3 | 27.8 | 36 | 37.4 | — | 33.2 | 29.5 | 21 | 28.1 | — | 29.3 | 21 | — | 29.7 | — | |
| GraphMDN (Oracle, Wide)Input=Ground-Truth 2D, Network=Wide, Selection=Oracle2020.10 | 28.9 | 34.5 | 28.2 | 30.2 | 31.5 | 32.3 | 28.6 | 35.7 | 43.3 | 31.9 | 38.5 | 32.1 | 25.2 | 33.3 | — | 31.8 | 27.8 | — | — | — | |
| PoseFormer (Zheng et al.)T (number of input frames)=81, Temporal information usage=true, Transformer-based architecture=true, Input=2D Ground Truth Keypoints2022.03 | 30 | 33.6 | 29.9 | 31 | 30.2 | 34.8 | 31.4 | 37.8 | 38.6 | 31.7 | 33.3 | 31.5 | 23.3 | 29 | — | 31.3 | 23.1 | — | — | — | |
| Zheng et al.input_source=GT, temporal_information=true2023.07 | 30 | 33.6 | 29.9 | 31 | 30.2 | 34.8 | 31.4 | 37.8 | 38.6 | — | 33.3 | 31.5 | 23.3 | 29 | — | 31.3 | 23.1 | — | 31.7 | — | |
| Zheng et al.Input Source=GT, Video-based=true2023.09 | 30 | 33.6 | 29.9 | 31 | 30.2 | 34.8 | 31.4 | 37.8 | 38.6 | — | 33.3 | 31.5 | 23.3 | 29 | — | 31.3 | 23.1 | — | 31.7 | — | |
| PoseFormerInput=Ground truth 2D pose, f=812021.03 | 30 | 33.6 | 29.9 | — | 30.2 | 34.8 | 31.4 | 37.8 | 38.6 | — | 33.3 | 31.5 | 23.3 | 29 | — | 31.3 | 23.1 | — | 31.7 | 31 | |
| Iskakov et al.Input Type=Multi-view, Extrinsic Camera Parameters=Not Given, Non-ep-free algorithm=true, Parameter perturbation=4%2021.05 | 30.2 | 37.2 | 32.7 | 33.2 | 33.8 | 29.7 | 43 | 49.4 | 67.6 | — | 43.7 | 33.1 | 27.2 | 42.1 | — | 38.4 | 29.3 | — | 38 | — | |
| DC-GCTInput Source=Ground Truth 2D, Evaluation Protocol=Protocol #12023.08 | 30.7 | 35.6 | 26.8 | 32 | 32.6 | 36 | 30 | 35.1 | 36.5 | — | 36.7 | 33.6 | 26.5 | 33.9 | — | 32.4 | 27.5 | — | 32.2 | — | |
| Ours (PoseNet)groundtruth info in inference time=true2019.07 | 31 | 30.6 | 39.9 | 35.5 | 34.8 | 30.2 | 32.1 | 35 | 43.8 | 35.7 | 37.6 | 30.1 | 24.6 | 35.7 | 29.3 | 34 | — | — | — | — | |
| MöbiusGCN#Param.=0.16M, 2D Input=GT2022.03 | 31.2 | 46.9 | 32.5 | 31.7 | 41.4 | 33.9 | 30.9 | 49.2 | 55.7 | 35.9 | 44.9 | 36.1 | 29.7 | 37.5 | — | 36.2 | 33.1 | — | — | — | |
| Zeng et al.Input Source=GT, Video-based=true2023.09 | 31.2 | 34.8 | 32.1 | 28.5 | 30.7 | 36.9 | 35.6 | 46.8 | 40.1 | — | 31.4 | 31 | 22.5 | 29.9 | — | 32 | 24.5 | — | 32.5 | — | |
| FLEXInput Type=Multi-view, Extrinsic Camera Parameters=Not Given, Exploit Temporal Information=true2021.05 | 32 | 23.6 | 24.3 | 26.7 | 30.4 | 25.1 | 32.9 | 29.5 | 32.5 | — | 35.7 | 26.5 | 26 | 34.7 | — | 30.2 | 27.7 | — | 32.6 | — | |
| Zhao et al.Input Source=Ground Truth 2D, Evaluation Protocol=Protocol #1, Venue/Year=CVPR 20222023.08 | 32 | 38 | 30.4 | 34.4 | 34.7 | 35.2 | 31.4 | 38 | 46.2 | — | 43.3 | 35.7 | 27.4 | 36.1 | — | 35.2 | 30.6 | — | 34.2 | — | |
| Lee et al.input_source=GT, temporal_information=true2023.07 | 32.1 | 36.6 | 34.3 | 37.8 | 44.5 | 40.9 | 36.2 | 44.1 | 45.6 | — | 49.9 | 35.9 | 37.6 | 30.3 | — | 38.4 | 35.5 | — | 35.3 | — | |
| GraphMDN (Mean)Input=Ground-Truth 2D, Network=Standard, Selection=Mean2020.10 | 32.2 | 39.5 | 33.5 | 36.7 | 38.2 | 39 | 36.7 | 44 | 52.9 | 37.2 | 48.5 | 40.3 | 30 | 39.4 | — | 38.7 | 32.1 | — | — | — | |
| Zhang et al.Input Source=Ground Truth 2D, Evaluation Protocol=Protocol #1, Venue/Year=IJCV'20232023.08 | 32.4 | 36.5 | 30.1 | 33.3 | 36.3 | 36.1 | 30.5 | 37.5 | 45.3 | — | 43.5 | 35.1 | 27.5 | 35.3 | — | 34.9 | 30.2 | — | 33.8 | — | |
| Ours (Full)groundtruth info in inference time=false2019.07 | 32.5 | 31.5 | 41.5 | 36.7 | 36.3 | 31.9 | 33.2 | 36.5 | 44.4 | 36.7 | 38.7 | 31.2 | 25.6 | 37.1 | 30.5 | 35.2 | — | — | — | — | |
| Li et al.Input Source=Ground Truth 2D, Evaluation Protocol=Protocol #1, Venue/Year=AAAI'20232023.08 | 32.9 | 38.3 | 28.3 | 33.8 | 34.9 | 37.2 | 30.7 | 34.5 | 39.7 | — | 38.7 | 34.7 | 26.1 | 34.3 | — | 33.8 | 28.9 | — | 33.9 | — | |
| Cai et al.Input Source=Ground Truth 2D, Evaluation Protocol=Protocol #1, Venue/Year=ICCV'20192023.08 | 33.4 | 39 | 33.8 | 37 | 38.1 | 39.5 | 37.3 | 43.2 | 46.2 | — | 47.3 | 38 | 30.4 | 38.6 | — | 38.1 | 32.1 | — | 37.7 | — | |
| MöbiusGCN#Param.=0.04M, 2D Input=GT2022.03 | 33.6 | 48.5 | 34.9 | 34.8 | 46 | 36.7 | 33.7 | 50.6 | 62.7 | 38.9 | 49.5 | 40.3 | 33.1 | 41.4 | — | 40 | 36.3 | — | — | — | |
| GraphMDN (Mean, Wide)Input=Ground-Truth 2D, Network=Wide, Selection=Mean2020.10 | 33.9 | 39.9 | 33 | 35.4 | 36.8 | 38.9 | 33 | 41 | 50 | 36.4 | 44.4 | 38.3 | 28.2 | 37.8 | — | 37.2 | 31.5 | — | — | — | |
| ZhouMulti-Hypothesis=false, Weakly Supervised=false2020.08 | 34.4 | 42.4 | 36.6 | 42.1 | 38.2 | 34.7 | 40.2 | 45.6 | 60.8 | 39 | 39.8 | 42.6 | 29.8 | 42 | — | 39.9 | 31.7 | — | — | — | |
| SemGCN (Wide)Input=Ground-Truth 2D, Network=Wide2020.10 | 34.4 | 41.6 | 32.9 | 37.5 | 39.2 | 41 | 34.1 | 43.6 | 52.3 | 37.6 | 47.5 | 40.7 | 30.5 | 38.6 | — | 39 | 32.9 | — | — | — | |
| GraphMDN (Highest, Wide)Input=Ground-Truth 2D, Network=Wide, Selection=Highest2020.10 | 34.5 | 40.5 | 33 | 35.7 | 37 | 39.1 | 33 | 41.2 | 50.2 | 36.6 | 44.8 | 38.6 | 28.4 | 38.2 | — | 37.5 | 31.8 | — | — | — | |
| Liu et al.T (number of input frames)=243, Temporal information usage=true, Input=2D Ground Truth Keypoints2022.03 | 34.5 | 37.1 | 33.6 | 34.2 | 32.9 | 39.6 | 35.8 | 40.7 | 41.4 | 33 | 37.1 | 33.8 | 26.6 | 33 | — | 34.7 | 26.9 | — | — | — | |
| Liu et al.input_source=GT, temporal_information=true2023.07 | 34.5 | 37.1 | 33.6 | 34.2 | 32.9 | 39.6 | 35.8 | 40.7 | 41.4 | — | 37.1 | 33.8 | 26.6 | 33 | — | 34.7 | 26.9 | — | 33 | — | |
| Liu et al.Input=Ground truth 2D pose, f=2432021.03 | 34.5 | 37.1 | 33.6 | — | 32.9 | 39.6 | 35.8 | 40.7 | 41.4 | — | 37.1 | 33.8 | 26.6 | 33 | — | 34.7 | 26.9 | — | 33 | 34.2 | |
| SemGCNInput=Ground-Truth 2D, Network=Standard2020.10 | 34.7 | 41.9 | 35.3 | 38.7 | 41 | 41.1 | 37.3 | 44.6 | 53.9 | 40.3 | 53.5 | 41.9 | 33.3 | 41.5 | — | 41 | 36 | — | — | — | |
| SRNetInput=Ground truth 2D pose2021.03 | 34.8 | 32.1 | 28.5 | — | 31.4 | 35.6 | 30.5 | 38.9 | 40.5 | — | 36.9 | 31 | 22.5 | 29.9 | — | 32 | 24.5 | — | 32.5 | 30.7 | |
| Pavllo et al.Input Source=GT, Video-based=true2023.09 | 35.2 | 40.2 | 32.7 | 35.7 | 38.2 | 40.6 | 36.1 | 48.8 | 47.3 | — | 45.5 | 39.7 | 27.8 | 38.7 | — | 37.8 | 29.5 | — | 37.8 | — | |
| Hossain et al.Input=Ground truth 2D pose2021.03 | 35.2 | 40.8 | 37.2 | — | 43.2 | 38.9 | 35.6 | 42.3 | 44.6 | — | 44 | 39.7 | 32.8 | 40.2 | — | 39.2 | 35.5 | — | 39.7 | 37.4 | |
| Hu et al.Input Type=Monocular, Exploit Temporal Information=true2021.05 | 35.5 | 41.3 | 36.6 | 39.1 | 39.2 | 39.9 | 37 | 51.9 | 63.3 | — | 49 | 41.4 | 29.8 | 40.3 | — | 41.1 | 28.9 | — | 40.9 | — | |
| GraphSH#Param.=3.7M, 2D Input=GT2022.03 | 35.8 | 38.1 | 31 | 35.3 | 43.2 | 31.7 | 38.4 | 45.5 | 35.4 | 36.7 | 37.3 | 36.8 | 30.7 | 27.9 | — | 35.8 | — | — | — | — | |
| Xu et al.Input Source=Ground Truth 2D, Evaluation Protocol=Protocol #1, Venue/Year=CVPR 20212023.08 | 35.8 | 38.1 | 31 | 35.3 | 35.8 | 37.3 | 31.7 | 38.4 | 45.5 | — | 43.2 | 36.7 | 27.9 | 36.8 | — | 35.8 | 30.7 | — | 35.4 | — | |
| Xu and Takanoinput_source=GT, temporal_information=false2023.07 | 35.8 | 38.1 | 31 | 35.3 | 35.8 | 37.3 | 31.7 | 38.4 | 45.5 | — | 43.2 | 36.7 | 27.9 | 36.8 | — | 35.8 | 30.7 | — | 35.4 | — | |
| Zeng et al.Input Source=Ground Truth 2D, Evaluation Protocol=Protocol #1, Venue/Year=ECCV'20202023.08 | 35.9 | 36.7 | 29.3 | 34.5 | 36 | 37.7 | 31.7 | 40.1 | 44.3 | — | 42.8 | 37.2 | 33.7 | 36.2 | — | 36.4 | 34 | — | 35.8 | — | |
| Cheng et al.Input Type=Monocular, Exploit Temporal Information=true2021.05 | 36.2 | 38.1 | 42.7 | 35.9 | 38.2 | 36.8 | 42 | 45.9 | 51.3 | — | 45.7 | 41.5 | 33.1 | 43.8 | — | 40.1 | 28.6 | — | 41.8 | — | |
| RTPCAInput Source=CPN, Video-based=true2023.09 | 36.2 | 38.9 | 36.7 | 37.3 | 41.7 | 36.6 | 39.5 | 51.3 | 57.2 | — | 48.3 | 39.5 | 27.3 | 40.6 | — | 40.1 | 27.3 | — | 41.4 | — | |
| Ci et al.input_source=GT, temporal_information=false2023.07 | 36.3 | 38.8 | 29.7 | 37.8 | 34.6 | 39.8 | 32.5 | 36.2 | 39.5 | — | 42.5 | 38.4 | 31.3 | 38.2 | — | 36.3 | 34.2 | — | 34.4 | — | |
| IVTf=5, transformer-based=true2022.08 | 36.5 | 40.1 | 38.4 | 40.7 | 42.6 | 30.1 | 43.4 | 46.1 | 58 | 40.2 | 42.8 | 37.1 | 32.1 | 40.8 | — | — | 33.5 | 40.2 | — | — | |
| Liu et al.#Param.=4.2M, 2D Input=GT2022.03 | 36.8 | 40.3 | 33 | 36.3 | 45 | 34.9 | 40.3 | 47.7 | 37.4 | 38.5 | 39.7 | 38.6 | 32 | 29.6 | — | 37.5 | — | — | — | — | |
| Liu et al.Input Source=Ground Truth 2D, Evaluation Protocol=Protocol #1, Venue/Year=ECCV'20202023.08 | 36.8 | 40.3 | 33 | 36.3 | 37.5 | 39.7 | 34.9 | 40.3 | 47.7 | — | 45 | 38.5 | 29.6 | 38.6 | — | 37.8 | 32 | — | 37.4 | — | |
| Sun [44]groundtruth info in inference time=true2019.07 | 36.9 | 36.2 | 40.6 | 40.4 | 41.9 | 34.9 | 35.7 | 50.1 | 59.4 | 40.4 | 44.9 | 39 | 30.8 | 39.8 | 36.7 | 40.6 | — | — | — | — | |
| Xu et al.input_source=CPN, temporal_information=true2023.07 | 37.4 | 43.5 | 42.7 | 42.7 | 46.6 | 41.3 | 45.1 | 52.7 | 60.2 | — | 59.7 | 43.1 | 33.7 | 47.7 | — | 45.6 | 37.1 | — | 45.8 | — | |
| Xu et al. (2020)Input Source=CPN, Video-based=true2023.09 | 37.4 | 43.5 | 42.7 | 42.7 | 46.6 | 41.3 | 45.1 | 52.7 | 60.2 | — | 59.7 | 43.1 | 33.7 | 47.7 | — | 45.6 | 37.1 | — | 45.8 | — | |
| Zhang et al.Input Source=CPN, Video-based=true2023.09 | 37.6 | 40.9 | 37.3 | 39.7 | 42.3 | 40.1 | 39.8 | 51.7 | 55 | — | 49.9 | 39.8 | 27.9 | 41 | — | 40.9 | 27.9 | — | 42.1 | — | |
| Martinez et al.Input Source=Ground Truth 2D, Evaluation Protocol=Protocol #1, Venue/Year=ICCV'20172023.08 | 37.7 | 44.4 | 40.3 | 42.1 | 48.2 | 44.4 | 42.1 | 54.6 | 58 | — | 54.9 | 46.4 | 36.4 | 47.6 | — | 45.5 | 40.4 | — | 45.1 | — | |
| Martinez et al.input_source=GT, temporal_information=false2023.07 | 37.7 | 44.4 | 40.3 | 42.1 | 48.2 | 44.4 | 42.1 | 54.6 | 58 | — | 54.9 | 46.4 | 36.4 | 47.6 | — | 45.5 | 40.4 | — | 45.1 | — | |
| SemGCN (GT)Input Source=2D Ground Truth (GT), Configuration=#12019.04 | 37.8 | 49.4 | 37.6 | 40.9 | 45.1 | 40.1 | 48.3 | 50.1 | 42.2 | 53.5 | 41.4 | 44.3 | 47.3 | 40.5 | — | 43.8 | 39 | — | — | — | |
| SemGCN (reported)Input=Ground-Truth 2D, Network=Standard2020.10 | 37.8 | 49.4 | 37.6 | 40.9 | 45.1 | 40.1 | 48.3 | 50.1 | 42.2 | 53.5 | 41.4 | 44.3 | 47.3 | 40.5 | — | 43.8 | 39 | — | — | — | |
| SemGCN#Param.=0.43M, 2D Input=GT2022.03 | 37.8 | 49.4 | 37.6 | 40.9 | 41.4 | 48.3 | 50.1 | 42.2 | 53.5 | 44.3 | 40.1 | 40.5 | 39 | 47.3 | — | 43.8 | — | — | — | — | |
| Zhao et al.input_source=GT, temporal_information=false2023.07 | 37.8 | 49.4 | 37.6 | 40.9 | 45.1 | 40.1 | 48.3 | 50.1 | 42.2 | — | 41.4 | 44.3 | 47.3 | 40.5 | — | 43.8 | 39 | — | 53.5 | — | |
| Hu et al.Input Source=CPN, Video-based=true2023.09 | 38 | 43.3 | 39.1 | 39.4 | 45.8 | 41.4 | 41.4 | 55.5 | 61.9 | — | 53.6 | 41.9 | 31.6 | 44.5 | — | 43.4 | 29.4 | — | 44.6 | — | |
| Fanggroundtruth info in inference time=true2019.07 | 38.2 | 41.7 | 43.7 | 44.9 | 48.5 | 40.2 | 38.2 | 54.5 | 64.4 | 47.2 | 55.3 | 44.3 | 36.7 | 47.3 | 41.7 | 45.7 | — | — | — | — | |
| Reddy et al.Input Type=Monocular, Exploit Temporal Information=true2021.05 | 38.4 | 46.8 | 44.3 | 43.2 | 44.8 | 52.9 | 36.7 | 45.3 | 54.5 | — | 48.3 | 44.4 | 46.2 | 41.9 | — | 44.6 | 39.9 | — | 63.4 | — | |
| Zhou et al.input_source=CPN, temporal_information=true2023.07 | 38.5 | 45.8 | 40.3 | 54.9 | 39.5 | 39.2 | 43.1 | 49.2 | 71.1 | — | 45.9 | 53.6 | 33.2 | 44.5 | — | 45.1 | 34.1 | — | 41 | — | |
| Shan et al.input_source=CPN, temporal_information=true2023.07 | 38.9 | 42.7 | 40.4 | 41.1 | 45.6 | 40.9 | 39.9 | 55.5 | 59.4 | — | 49.7 | 42.2 | 29.4 | 42.7 | — | 42.8 | 29.4 | — | 44.9 | — | |
| MHFormerBounding box source=Ground truth2026.01 | 39.17 | 43.06 | 40.07 | 40.94 | 44.95 | 40.63 | 41.29 | 53.56 | 60.37 | 43.71 | 51.16 | 42.12 | 29.76 | 43.84 | — | 42.95 | 30.59 | — | — | — | |
| Li et al.input_source=CPN, temporal_information=true2023.07 | 39.2 | 43.1 | 40.1 | 40.9 | 44.9 | 40.6 | 41.3 | 53.5 | 60.3 | — | 51.2 | 41.1 | 29.8 | 43.8 | — | 43 | 30.6 | — | 43.7 | — | |
| Li et al.Input Source=CPN, Video-based=true2023.09 | 39.2 | 43.1 | 40.1 | 40.9 | 44.9 | 40.6 | 41.3 | 53.5 | 60.3 | — | 51.2 | 41.1 | 29.8 | 43.8 | — | 43 | 30.6 | — | 43.7 | — | |
| Kadkhodamohammadi and PadoyInput Type=Multi-view, Extrinsic Camera Parameters=Given2021.05 | 39.4 | 46.9 | 41 | 42.7 | 53.6 | 41.4 | 50 | 59.9 | 78.8 | — | 54.8 | 44.2 | 39.3 | 41.3 | — | 49.1 | 41 | — | 42.8 | — | |
| Martinezgroundtruth info in inference time=true2019.07 | 39.5 | 43.2 | 46.4 | 47 | 51 | 41.4 | 40.6 | 56.5 | 69.4 | 49.2 | 56 | 45 | 38 | 49.5 | 43.1 | 47.7 | — | — | — | — | |
| LlopartInput Type=Monocular, Exploit Temporal Information=true2021.05 | 39.5 | 41.7 | 45.6 | 42 | 43 | 42.5 | 41.7 | 55.2 | 62.3 | — | 53.9 | 42.9 | 31.8 | 45.3 | — | 44.8 | 31.8 | — | 44.9 | — | |
| Li et al.Input Type=Monocular, Exploit Temporal Information=true2021.05 | 39.9 | 43.4 | 40 | 40.9 | 42.4 | 42.1 | 39.8 | 55.8 | 61.6 | — | 50.6 | 43.3 | 29.9 | 44.9 | — | 43.6 | 30.3 | — | 44.9 | — | |
| LeeTemporal models=true, Ground-truth bounding box=true2019.10 | 40.2 | 49.2 | 47.8 | 52.6 | 50.1 | 50.2 | 43 | 55.8 | 73.9 | 54.1 | 75 | 55.6 | 43.3 | 58.2 | — | 52.8 | 43.3 | — | — | — | |
| Lee et al.input_source=CPN, temporal_information=true2023.07 | 40.2 | 49.2 | 47.8 | 52.6 | 50.1 | 50.2 | 43 | 55.8 | 73.9 | — | 75 | 55.6 | 43.3 | 58.2 | — | 52.8 | 43.3 | — | 54.1 | — | |
| PoseFormerV2Bounding box source=Ground truth2026.01 | 41.26 | 45.45 | 41.47 | 44.04 | 46.67 | 42.63 | 42.64 | 55.19 | 64.62 | 45.71 | 53.75 | 42.89 | 32.34 | 45.76 | — | 45.15 | 32.89 | — | — | — | |
| Wang et al.f=96, transformer-based=false2022.08 | 41.3 | 43.9 | 44 | 42.2 | 48 | 42.2 | 43.2 | 57.3 | 61.3 | 47 | 57.1 | 43.5 | 32.6 | 47 | — | — | 31.8 | 45.6 | — | — | |
| GLA-GCNinput_source=CPN, temporal_information=true, T=2432023.07 | 41.3 | 44.3 | 40.8 | 41.8 | 45.9 | 42.1 | 41.5 | 57.8 | 62.9 | — | 54.1 | 42.8 | 29.4 | 45.9 | — | 44.4 | 29.9 | — | 45 | — | |
| Wang et al.Input Source=CPN, Video-based=true2023.09 | 41.3 | 43.9 | 44 | 42.2 | 48 | 42.2 | 43.2 | 57.3 | 61.3 | — | 57.1 | 43.5 | 32.6 | 47 | — | 45.6 | 31.8 | — | 47 | — | |
| Anatomy3DBounding box source=Ground truth2026.01 | 41.49 | 43.8 | 39.77 | 43.12 | 46.18 | 42.16 | 41.85 | 54.15 | 60.7 | 45.49 | 52.54 | 41.56 | 31.41 | 46.04 | — | 44.18 | 32.45 | — | — | — | |
| Zheng et al.f=81, transformer-based=true2022.08 | 41.5 | 44.8 | 39.8 | 42.5 | 46.5 | 42.1 | 42 | 53.3 | 60.7 | 45.5 | 51.6 | 43.3 | 31.8 | 46.1 | — | — | 32.2 | 44.3 | — | — | |
| Zheng et al.Input Source=CPN, Video-based=true2023.09 | 41.5 | 44.8 | 39.8 | 42.5 | 46.5 | 42.1 | 42 | 53.3 | 60.7 | — | 51.6 | 43.3 | 31.8 | 46.1 | — | 44.3 | 32.2 | — | 45.5 | — | |
| Liu et al.f=243, transformer-based=false, backbone=CVPR 202022.08 | 41.8 | 44.8 | 41.1 | 44.9 | 47.4 | 43.4 | 42.2 | 56.2 | 63.6 | 45.3 | 54.1 | 43.5 | 31.3 | 45.3 | — | — | 32.2 | 45.1 | — | — | |
| Liu et al.input_source=CPN, temporal_information=true2023.07 | 41.8 | 44.8 | 41.1 | 44.9 | 47.4 | 43.4 | 42.2 | 56.2 | 63.6 | — | 54.1 | 43.5 | 31.3 | 45.3 | — | 45.1 | 32.2 | — | 45.3 | — | |
| Liu et al.Input Source=CPN, Video-based=true2023.09 | 41.8 | 44.8 | 41.1 | 44.9 | 47.4 | 43.4 | 42.2 | 56.2 | 63.6 | — | 54.1 | 43.5 | 31.3 | 45.3 | — | 45.1 | 32.2 | — | 45.3 | — | |
| Sun [43]groundtruth info in inference time=true2019.07 | 42.1 | 44.3 | 45 | 45.4 | 51.5 | 43.2 | 41.3 | 59.3 | 73.3 | 51 | 53 | 44 | 38.3 | 48 | 44.8 | 48.3 | — | — | — | — | |
| Chen et al.f=81, transformer-based=false2022.08 | 42.1 | 43.8 | 41 | 43.8 | 46.1 | 42.4 | 43.1 | 53.9 | 60.5 | 45.7 | 53.5 | 42.1 | 32.2 | 46.2 | — | — | 33.8 | 44.6 | — | — | |
| Lin et al.f=50, transformer-based=false2022.08 | 42.5 | 44.8 | 42.6 | 44.2 | 48.5 | 52.6 | 41.4 | 56.5 | 64.5 | 47.4 | 57.1 | 43 | 33 | 48.1 | — | — | 35.1 | 46.6 | — | — | |
| Liu et al.f=243, transformer-based=true, backbone=ICRA 212022.08 | 43.3 | 46.1 | 40.9 | 44.6 | 46.6 | 44.1 | 42.9 | 55.3 | 57.9 | 45.8 | 54 | 43.4 | 30.4 | 47.3 | — | — | 30.3 | 44.9 | — | — | |
| Tome et al.Input Type=Multi-view, Extrinsic Camera Parameters=Given, Extra Training Data=true2021.05 | 43.3 | 49.6 | 42 | 48.8 | 51.1 | 40.3 | 43.3 | 66 | 95.2 | — | 64.3 | 52.2 | 48.3 | 51.1 | — | 52.8 | 46.3 | — | 50.2 | — |