Table 3.
RGB-D-based pose estimation methods and results.
| Methods | Years | Input | LM | LM-O | YCB-V |
|---|---|---|---|---|---|
| Li et al. [102] | 2018 | RGB-D | - | - | 94.3 |
| DenseFusion [103] | 2019 | RGB-D | 94.3 | - | 91.2 |
| Morefusion [5] | 2020 | RGB-D | - | - | 91.0 |
| PVN3D [105] | 2020 | RGB-D | 99.4 | 70.2 | 91.8 |
| G2L-Net [109] | 2020 | RGB-D | 98.7 | - | 92.4 |
| PR-GCN [114] | 2020 | RGB-D | 99.6 | 65.0 | 95.8 |
| FFB6D [104] | 2021 | RGB-D | 99.7 | 66.2 | 92.7 |
| Uni6d [108] | 2022 | RGB-D | - | - | 88.8 |
| E2EK [106] | 2022 | RGB-D | 99.8 | 75.3 | 94.4 |
| RCVPose [115] | 2022 | RGB-D | 99.4 | 70.2 | 95.2 |
| Deepfusion [107] | 2023 | RGB-D | 99.8 | 77.7 | 94.4 |