6D Object Pose Estimation on OccludedLINEMOD (test)
82.5ADD(S)OVE6D
Evaluation results
| Method | Links | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OVE6DGeneralization=Universal Objects, Train Data=Syn. Only, Input=D, ICP=true, Mask=Ground Truth, ICP for all proposals=true2022.03 | 82.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OVE6DGeneralization=Universal Objects, Train Data=Syn. Only, Input=D, ICP=true, Mask=Ground Truth2022.03 | 80 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PVNetGeneralization=Single Trained Object, Train Data=Real (+Syn.), Input=RGBD, ICP=true2022.03 | 79 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PoseCNNGeneralization=Multi-Trained Objects, Train Data=Real + Syn., Input=RGBD, ICP=true2022.03 | 78 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OVE6DGeneralization=Universal Objects, Train Data=Syn. Only, Input=D, ICP=true, Mask=Mask R-CNN, ICP for all proposals=true2022.03 | 72.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RCVPoseMode=RGB-D, refinement=ICP2021.04 | 71.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OVE6DGeneralization=Universal Objects, Train Data=Syn. Only, Input=D, ICP=false, Mask=Ground Truth2022.03 | 70.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OVE6DGeneralization=Universal Objects, Train Data=Syn. Only, Input=D, ICP=true, Mask=Mask R-CNN2022.03 | 70.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RCVPoseMode=RGB-D2021.04 | 70.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FFB6DGeneralization=Single Trained Object, Train Data=Real (+Syn.), Input=RGBD, ICP=false2022.03 | 66.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FFB6DMode=RGB-D2021.04 | 66.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CloudAAE(GT)Generalization=Multi-Trained Objects, Train Data=Real + Syn., Input=D, ICP=true2022.03 | 66.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PR-GCNGeneralization=Multi-Trained Objects, Train Data=Real + Syn., Input=RGBD, ICP=false2022.03 | 65 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PR-GCNMode=RGB-D2021.04 | 65 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PVN3DGeneralization=Single Trained Object, Train Data=Real (+Syn.), Input=RGBD, ICP=false2022.03 | 63.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CloudAAE(GT)Generalization=Multi-Trained Objects, Train Data=Syn. Only, Input=D, ICP=true2022.03 | 63.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PVN3DMode=RGB-D2021.04 | 63.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SO-PoseMode=RGB2021.04 | 62.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SO-PoseRefinement=w/o Refinement, P.E.=1, Training Data=real + syn2021.08 | 62.3 | 48.4 | 85.8 | 3,270 | 77.4 | 48.9 | 52.4 | 78.3 | 75.3 | — | — | — | — | — | |
| GDR-NetMode=RGB2021.04 | 62.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CloudAAE(GT)Generalization=Multi-Trained Objects, Train Data=Real + Syn., Input=D, ICP=false2022.03 | 58.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PPRNMode=RGB2021.04 | 58.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CloudAAE(GT)Generalization=Multi-Trained Objects, Train Data=Syn. Only, Input=D, ICP=false2022.03 | 57.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OVE6DGeneralization=Universal Objects, Train Data=Syn. Only, Input=D, ICP=false, Mask=Mask R-CNN2022.03 | 56.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SO-PoseRefinement=w/o Refinement, P.E.=1, Training Data=real + pbr + syn2021.08 | 56.1 | 44.9 | 79.7 | 3,060 | 67.8 | 40 | 49.8 | 73.7 | 62.7 | — | — | — | — | — | |
| DeepIMRefinement=w/ Refinement, P.E.=1, Training Data=real + syn2021.08 | 55.5 | 59.2 | 63.5 | 2,620 | 55.6 | 52.4 | 63 | 71.7 | 52.5 | — | — | — | — | — | |
| SO-PoseRefinement=w/o Refinement, P.E.=1, Training Data=real + pbr2021.08 | 54.3 | 46.3 | 81.1 | 1,870 | 71.3 | 43.9 | 46.6 | 63.3 | 62.9 | — | — | — | — | — | |
| HybridPoseRefinement=w/o Refinement, P.E.=M, Training Data=real + syn2021.08 | 47.5 | 20.9 | 75.3 | 2,490 | 70.2 | 27.9 | 52.4 | 53.8 | 54.2 | — | — | — | — | — | |
| GDR-NetRefinement=w/o Refinement, P.E.=1, Training Data=real + syn2021.08 | 47.4 | 41.3 | 71.1 | 1,820 | 54.6 | 41.7 | 40.2 | 59.5 | 52.6 | — | — | — | — | — | |
| DPODMode=RGB + D ref, refinement=true2021.04 | 47.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DPODRefinement=w/ Refinement, P.E.=1, Training Data=real + syn2021.08 | 47.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CloudPose (GT)Generalization=Multi-Trained Objects, Train Data=Syn. Only, Input=D, ICP=true2022.03 | 44.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| S.StageRefinement=w/o Refinement, P.E.=M, Training Data=real + syn2021.08 | 43.3 | 19.2 | 65.1 | 1,890 | 69 | 25.3 | 52 | 51.4 | 45.6 | — | — | — | — | — | |
| PVNetGeneralization=Single Trained Object, Train Data=Real (+Syn.), Input=RGB, ICP=false2022.03 | 42.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PVNetMode=RGB2021.04 | 40.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PVNetRefinement=w/o Refinement, P.E.=M, Training Data=real + syn2021.08 | 40.8 | 15.8 | 63.3 | 1,670 | 65.7 | 25.2 | 50.2 | 49.6 | 36.1 | — | — | — | — | — | |
| DPODMode=RGB2021.04 | 32.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Pix2PoseMode=RGB2021.04 | 32 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Hu et al.Mode=RGB2021.04 | 30.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SSD6DMode=RGB + D ref, refinement=true2021.04 | 27.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OberwegerMode=RGB2021.04 | 27.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PoseCNNGeneralization=Multi-Trained Objects, Train Data=Real + Syn., Input=RGB, ICP=false2022.03 | 24.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PoseCNNMode=RGB + D ref, refinement=true2021.04 | 24.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PoseCNNRefinement=w/o Refinement, P.E.=1, Training Data=real + syn2021.08 | 24.9 | 9.6 | 45.2 | 90 | 41.4 | 19.6 | 22 | 38.5 | 22.1 | — | — | — | — | — | |
| YOLO6DMode=RGB2021.04 | 6.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| BB82019.09 | — | — | — | — | — | — | — | — | — | — | — | 7.6 | — | — | |
| Brachmann et al.Input Modality=RGB-D2017.11 | — | 53.1 | 79.9 | 28.2 | 82 | 64.3 | 9 | 44.5 | 91.6 | 56.6 | — | — | — | — | |
| CullNetbias correction=true2019.09 | — | — | — | — | — | — | — | — | — | — | — | 38.59 | 24.48 | — | |
| DeepIM2023.06 | — | — | — | — | — | — | — | — | — | — | — | — | — | 55.5 | |
| DPOD2020.01 | — | — | — | — | — | — | — | — | — | 47.3 | — | — | — | — | |
| DPODRefinement=Without2021.10 | — | — | — | — | — | — | — | — | — | 32.8 | — | — | — | — | |
| DPOD2022.03 | — | — | — | — | — | — | — | — | — | 47.3 | — | — | — | — | |
| DPOD+Refinement=With2021.10 | — | — | — | — | — | — | — | — | — | 47.3 | — | — | — | — | |
| GDRRefinement=Without2021.10 | — | 39.3 | 79.2 | 23.5 | 71.3 | 44.4 | 58.2 | 49.3 | 58.7 | 53 | — | — | — | — | |
| GDR-Net2022.03 | — | 39.3 | 79.2 | 2,350 | 71.3 | 44.4 | 58.2 | 49.3 | 58.7 | 53 | — | — | — | — | |
| Hinterstoisser et al.Input Modality=RGB-D2017.11 | — | 81.4 | 94.7 | 55.2 | 86 | 79.7 | 65.5 | 52.1 | 95.5 | 76.3 | — | — | — | — | |
| HMRefinement=Without2021.10 | — | 15.3 | 44.7 | 9.33 | 55.4 | 19.6 | 23 | 41.4 | 20.4 | 28.64 | — | — | — | — | |
| HuRefinement=Without2021.10 | — | 12.1 | 39.9 | 8.2 | 45.2 | 17.2 | 22.1 | 35.8 | 36 | 27.06 | — | — | — | — | |
| Hu et al.2020.01 | — | 17.6 | 53.9 | 330 | 62.4 | 19.2 | 25.9 | 39.6 | 21.3 | 27 | — | — | — | — | |
| Hu2Refinement=Without2021.10 | — | 19.2 | 65.1 | 18.9 | 69 | 25.3 | 52 | 51.4 | 45.6 | 43.3 | — | — | — | — | |
| Hybrid2021.04 | — | 20.9 | 75.3 | 24.9 | 70.2 | 27.9 | 52.4 | 53.8 | 54.2 | 47.5 | — | — | — | — | |
| HybridPose2020.01 | — | 20.9 | 75.3 | 2,490 | 70.2 | 27.9 | 52.4 | 53.8 | 54.2 | 47.5 | — | — | — | — | |
| HybridPoseRefinement=With2021.10 | — | 20.9 | 75.3 | 24.9 | 70.2 | 27.9 | 52.4 | 53.8 | 54.2 | 47.45 | — | — | — | — | |
| HybridPose2022.03 | — | 20.9 | 75.3 | 2,490 | 70.2 | 27.9 | 52.4 | 53.8 | 54.2 | 47.5 | — | — | — | — | |
| Jafari2019.09 | — | — | — | — | — | — | — | — | — | — | — | 20.8 | — | — | |
| Krull et al.Input Modality=RGB-D2017.11 | — | 68 | 87.9 | 50.6 | 91.2 | 64.7 | 41.5 | 65.3 | 92.9 | 70.3 | — | — | — | — | |
| Michel et al.Input Modality=RGB-D2017.11 | — | 80.7 | 88.5 | 57.8 | 94.7 | 74.4 | 47.6 | 73.8 | 96.3 | 76.7 | — | — | — | — | |
| OberwegerObject=ape2018.12 | — | — | — | — | — | — | — | — | — | — | 69.6 | — | — | — | |
| OberwegerObject=can2018.12 | — | — | — | — | — | — | — | — | — | — | 82.6 | — | — | — | |
| OberwegerObject=cat2018.12 | — | — | — | — | — | — | — | — | — | — | 65.1 | — | — | — | |
| OberwegerObject=duck2018.12 | — | — | — | — | — | — | — | — | — | — | 61.4 | — | — | — | |
| OberwegerObject=driller2018.12 | — | — | — | — | — | — | — | — | — | — | 73.8 | — | — | — | |
| OberwegerObject=eggbox2018.12 | — | — | — | — | — | — | — | — | — | — | 13.1 | — | — | — | |
| OberwegerObject=glue2018.12 | — | — | — | — | — | — | — | — | — | — | 54.9 | — | — | — | |
| OberwegerObject=holepuncher2018.12 | — | — | — | — | — | — | — | — | — | — | 66.4 | — | — | — | |
| OberwegerObject=average2018.12 | — | — | — | — | — | — | — | — | — | — | 60.9 | — | — | — | |
| Oberweger2019.08 | — | 17.6 | 53.9 | 3.31 | 62.4 | 19.2 | 25.9 | 39.6 | 21.3 | 30.4 | — | — | — | — | |
| Oberweger et al.2020.01 | — | 12.1 | 39.9 | 820 | 45.2 | 17.2 | 22.1 | 35.8 | 36 | 27 | — | — | — | — | |
| PFA2023.06 | — | — | — | — | — | — | — | — | — | — | — | — | — | 65.3 | |
| Pix2Pose2019.08 | — | 22 | 44.7 | 22.7 | 44.7 | 15 | 25.2 | 32.4 | 49.5 | 32 | — | — | — | — | |
| Pix2Pose2020.01 | — | 22 | 44.7 | 2,270 | 44.7 | 15 | 25.2 | 32.4 | 49.5 | 32 | — | — | — | — | |
| Pix2PoseRefinement=Without2021.10 | — | 22 | 44.7 | 22.7 | 44.7 | 15 | 25.2 | 32.4 | 49.5 | 32.03 | — | — | — | — | |
| PoseCNNObject=ape2018.12 | — | — | — | — | — | — | — | — | — | — | 34.6 | — | — | — | |
| PoseCNNObject=can2018.12 | — | — | — | — | — | — | — | — | — | — | 15.1 | — | — | — | |
| PoseCNNObject=cat2018.12 | — | — | — | — | — | — | — | — | — | — | 10.4 | — | — | — | |
| PoseCNNObject=duck2018.12 | — | — | — | — | — | — | — | — | — | — | 31.8 | — | — | — | |
| PoseCNNObject=driller2018.12 | — | — | — | — | — | — | — | — | — | — | 7.4 | — | — | — | |
| PoseCNNObject=eggbox2018.12 | — | — | — | — | — | — | — | — | — | — | 1.9 | — | — | — | |
| PoseCNNObject=glue2018.12 | — | — | — | — | — | — | — | — | — | — | 13.8 | — | — | — | |
| PoseCNNObject=holepuncher2018.12 | — | — | — | — | — | — | — | — | — | — | 23.1 | — | — | — | |
| PoseCNNObject=average2018.12 | — | — | — | — | — | — | — | — | — | — | 17.2 | — | — | — | |
| PoseCNN2019.09 | — | — | — | — | — | — | — | — | — | — | — | 17.2 | 24.9 | — | |
| PoseCNNtraining=synthetic and real images2019.08 | — | 9.6 | 45.2 | 0.93 | 41.4 | 19.6 | 22 | 38.5 | 22.1 | 24.9 | — | — | — | — | |
| PoseCNN2020.01 | — | 9.6 | 45.2 | 93 | 41.4 | 19.6 | 22 | 38.5 | 22.1 | 24.9 | — | — | — | — | |
| PoseCNN2022.03 | — | 9.6 | 45.2 | 93 | 41.4 | 19.6 | 22 | 38.5 | 22.1 | 24.9 | — | — | — | — | |
| PoseCNN2023.06 | — | — | — | — | — | — | — | — | — | — | — | — | — | 24.9 | |
| PoseCNN ColorInput Modality=Color Only2017.11 | — | 9.6 | 45.2 | 0.93 | 41.4 | 19.6 | 22 | 38.5 | 22.1 | 24.9 | — | — | — | — | |
| PoseCNN+ICPInput Modality=RGB-D, Refinement=ICP2017.11 | — | 76.2 | 87.4 | 52.2 | 90.3 | 77.7 | 72.2 | 76.7 | 91.4 | 78 | — | — | — | — | |
| PVNetObject=ape2018.12 | — | — | — | — | — | — | — | — | — | — | 69.14 | — | — | — | |
| PVNetObject=can2018.12 | — | — | — | — | — | — | — | — | — | — | 86.09 | — | — | — |