3D Semantic Scene Completion on SemanticKITTI 51.2m range (val)
44.15IoUVoxFormer-T
Evaluation Results
| Method | Links | |||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| VoxFormer-TTemporal Information=Current + Previous 4 images, Input Modality=Camera2023.02 | 44.15 | 62.06 | 60.47 | 13.35 | 26.54 | 1.28 | 0.56 | 7.26 | 7.81 | 1.93 | 1.97 | 0 | 53.57 | 19.69 | 26.52 | 0.42 | 19.54 | 7.31 | 26.1 | 6.1 | 33.06 | 9.15 | 4.14 | |
| VoxFormer-STemporal Information=Current image only, Input Modality=Camera2023.02 | 44.02 | 62.32 | 59.99 | 12.35 | 25.79 | 0.59 | 0.51 | 5.63 | 3.77 | 1.78 | 3.32 | 0 | 54.76 | 15.5 | 26.35 | 0.7 | 17.65 | 7.64 | 24.39 | 5.08 | 29.96 | 7.11 | 4.18 | |
| SSCNet*Input Modality=RGB-inferred (pseudo LiDAR)2023.02 | 40.93 | 48.77 | 71.8 | 10.27 | 22.32 | 0 | 0 | 4.69 | 2.43 | 0 | 0.01 | 0 | 51.28 | 9.07 | 22.38 | 0.02 | 15.2 | 3.57 | 22.24 | 4.33 | 31.21 | 4.83 | 1.49 | |
| LMSCNet*Input Modality=RGB-inferred (pseudo LiDAR)2023.02 | 38.36 | 77.6 | 43.13 | 9.94 | 23.62 | 0 | 0 | 1.69 | 0 | 0 | 0 | 0 | 54.9 | 9.89 | 25.43 | 0 | 14.55 | 3.27 | 20.19 | 1.06 | 32.3 | 2.04 | 0 | |
| MonoSceneInput Modality=Camera2023.02 | 36.8 | 52.19 | 55.5 | 11.3 | 23.29 | 0.28 | 0.59 | 9.29 | 2.63 | 2 | 1.07 | 0 | 55.89 | 14.75 | 26.5 | 1.63 | 13.55 | 6.6 | 17.98 | 2.44 | 29.84 | 3.91 | 2.43 |