Orientation estimation on T-LESS SiSo
14ARCSundermeyer et al. (2020)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Sundermeyer et al. (2020)Training=RGB: R,S, Test=RGBD, Scope=dataset, Model=CAD, Symm.=network2026.04 | 14 | 62.3 | 62.4 | 0.531 | |
| Hodaň et al. (2015)Training=templates, Test=RGBD, Scope=unknown, Model=CAD, Symm.=evaluation2026.04 | 17.1 | 70.1 | 66.2 | 80.1 | |
| Drost et al. (2010)Training=templates, Test=D, Scope=unknown, Model=reconstr., Symm.=repres.2026.04 | 17.9 | 70.3 | 64.2 | 9.2 | |
| ZTE PPF (2022)Training=templates, Test=D, Scope=unknown, Model=reconstr., Symm.=unknown2026.04 | 21.9 | 79.3 | 73.8 | 0.846 | |
| Vidal et al. (2018)Training=templates, Test=D, Scope=unknown, Model=reconstr., Symm.=repres.2026.04 | 22.6 | 76 | 74 | 7.06 | |
| Castro and Kim (2023)Training=RGB: S, Test=RGB, Scope=dataset, Model=CAD, Symm.=loss2026.04 | 24.8 | 73.9 | 83.7 | 0.059 | |
| ModalOcc.depth (2024)Training=RGBD: S, Test=D, Scope=unknown, Model=CAD, Symm.=unknown2026.04 | 27.3 | 81.8 | 83.4 | 4.72 | |
| SARR–RGBTraining=RGB: R, Test=RGB, Scope=symmetry, Model=✗, Symm.=repres.2026.04 | 29.1 | 42.4 | 49.9 | 0.077 | |
| Cai et al. (2022)Training=RGB: S, Test=RGB, Scope=dataset, Model=✗, Symm.=network2026.04 | 30.2 | 75.1 | 85.9 | 50.8 | |
| ModalOcc.rgbd (2024)Training=RGBD: S, Test=RGBD, Scope=unknown, Model=CAD, Symm.=unknown2026.04 | 30.8 | 79.8 | 90.4 | 7.24 | |
| Liu et al. (2025)Training=RGBD: R,S, Test=RGBD, Scope=object, Model=reconstr., Symm.=loss2026.04 | 32 | 89.4 | 91.1 | 2.48 | |
| ModalOcc.rgb (2024)Training=RGBD: S, Test=RGB, Scope=unknown, Model=CAD, Symm.=unknown2026.04 | 32.6 | 69.1 | 90 | 7.75 | |
| SARR–RGBTraining=RGB: R, Test=RGB, Scope=object, Model=✗, Symm.=repres.2026.04 | 32.6 | 46.8 | 58.2 | 0.077 | |
| Wang et al. (2021)Training=RGBD: R,S, Test=RGBD, Scope=object, Model=CAD, Symm.=loss2026.04 | 33.5 | 90.4 | 90.7 | 6.63 | |
| Liu et al. (2025)Training=RGB: R,S, Test=RGB, Scope=object, Model=reconstr., Symm.=loss2026.04 | 33.7 | 89.1 | 90.6 | 0.214 | |
| Su et al. (2022)Training=RGBD: R,S, Test=RGBD, Scope=object, Model=reconstr., Symm.=data2026.04 | 38.5 | 85.6 | 91.2 | 2.62 | |
| SARR–DepthTraining=D: R, Test=D, Scope=object, Model=✗, Symm.=repres.2026.04 | 45 | 48 | 68.4 | 0.077 | |
| SARR–RGBTraining=RGB: R, Test=RGB, Scope=dataset*, Model=✗, Symm.=repres.2026.04 | 45.8 | 52.1 | 69.1 | 0.08 | |
| SARR–DepthTraining=D: R, Test=D, Scope=dataset*, Model=✗, Symm.=repres.2026.04 | 46.7 | 54.9 | 68.8 | 0.077 | |
| SARR–RGBTraining=RGB: R, Test=RGB, Scope=dataset, Model=✗, Symm.=repres.2026.04 | 47.3 | 49 | 69.3 | 0.078 | |
| SARR–DepthTraining=D: R, Test=D, Scope=symmetry, Model=✗, Symm.=repres.2026.04 | 47.5 | 56.3 | 70 | 0.078 | |
| SARR–DepthTraining=D: R, Test=D, Scope=dataset, Model=✗, Symm.=repres.2026.04 | 48 | 50.3 | 69.8 | 0.074 |