3D Semantic Occupancy Prediction on SemanticKITTI (val)
84.65mIoUSemCity
Evaluation results
| Method | Links | |||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SemCityRepresentation=triplane2026.06 | 84.65 | 84.84 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SemCityRepresentation=BEV2026.06 | 77.84 | 80.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EditSSC2026.06 | 72.2 | 81.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SemCityRepresentation=BEV VQ-VAE2026.06 | 68.35 | 80.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DRVRModality=L2026.06 | 30.4 | 60.3 | — | 73.5 | 49.4 | 38.8 | 0.1 | 41.1 | 51 | 61.4 | 14.1 | 17.2 | 31.6 | 43.9 | 27 | 47.1 | 7.9 | 3.3 | 0 | 20.5 | 32.6 | 17.1 | — | |
| OCC-ExoskeletonModality=L2026.06 | 26.9 | 59.7 | — | 74 | 46.3 | 25.4 | 1.9 | 41.7 | 47.2 | 34.8 | 13.4 | 17.1 | 20 | 43.3 | 23.9 | 50.1 | 8 | 1.9 | 0 | 16.8 | 29.2 | 15.3 | — | |
| VPNetModality=L2026.06 | 26.1 | 59.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SSC-RSModality=L2026.06 | 24.8 | 58.6 | — | 73.8 | 45.3 | 26.6 | 2.1 | 41 | 46.8 | 41.5 | 1.5 | 6.9 | 19.8 | 42.6 | 22.2 | 50.6 | 6.2 | 1.5 | 0 | 15.8 | 17.9 | 4.6 | — | |
| SSA-SCModality=L2026.06 | 24.5 | 58.3 | — | 72.8 | 44.3 | 21.1 | 4.1 | 41.5 | 47 | 39.6 | 9.2 | 7.4 | 19.1 | 41.9 | 22 | 49.4 | 6.3 | 3.2 | 0 | 15.2 | 17.8 | 4.4 | — | |
| Local-DIFsModality=L2026.06 | 24 | 57.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| JS3C-NetModality=L2026.06 | 24 | 57 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| UDNetModality=L2026.06 | 20.7 | 58.9 | — | 67 | 37.2 | 20.3 | 2.2 | 36 | 42.1 | 25.7 | 1.8 | 2.3 | 11.2 | 40.1 | 18.3 | 45.8 | 2.5 | 1.2 | 0 | 11.9 | 23 | 3.8 | — | |
| ProOOD (+VoxDet)base_method=VoxDet, evaluation_mode=single-frame2026.04 | 18.86 | — | — | 67.18 | 34.87 | 23.64 | 0.59 | 24.54 | 34.09 | 30.6 | 3.75 | 5.95 | 15.57 | 29.18 | 9.67 | 41.25 | 3.75 | 3.95 | 0 | 10.59 | 11.62 | 7.57 | 8.46 | |
| VoxDetevaluation_mode=single-frame, implementation=reproduced2026.04 | 18.08 | — | — | 66.5 | 34.47 | 19.83 | 0.36 | 24.77 | 34.63 | 24.74 | 3.79 | 5.41 | 13.23 | 28.96 | 9.01 | 41.57 | 3.57 | 2.89 | 0 | 10.89 | 11.71 | 7.18 | 7.44 | |
| VisHall3DModality=C2026.06 | 17.1 | 46.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CGFormerModality=C2026.06 | 16.9 | 46 | — | 65.5 | 32.3 | 20.8 | 0.2 | 23.5 | 34.3 | 19.4 | 4.6 | 2.7 | 7.7 | 26.9 | 8.8 | 39.5 | 2.4 | 4.1 | 0 | 9.2 | 10.7 | 7.8 | — | |
| LMSCNet-SSModality=L2026.06 | 16.8 | 54.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ProOOD (+CGFormer)base_method=CGFormer, evaluation_mode=single-frame2026.04 | 16.24 | — | — | 66.87 | 33.61 | 24.63 | 0.34 | 22.61 | 33.78 | 1.42 | 5.5 | 3.41 | 5.94 | 27.01 | 8.18 | 38.51 | 2.87 | 5.33 | 0 | 10.7 | 10.62 | 7.22 | 4.62 | |
| CGFormerevaluation_mode=single-frame, implementation=reproduced2026.04 | 15.64 | — | — | 64.79 | 32.11 | 21.8 | 0.02 | 23.19 | 33.94 | 1.5 | 2.9 | 4.14 | 7.21 | 27.24 | 7.56 | 38.16 | 1.75 | 3.06 | 0 | 9.9 | 10.71 | 7.12 | 4.18 | |
| ProOOD (+SGN)base_method=SGN, evaluation_mode=single-frame2026.04 | 14.97 | — | — | 59.35 | 27.94 | 20.32 | 0.08 | 21.6 | 31.5 | 14.54 | 0.82 | 0.91 | 5.67 | 26.06 | 8.7 | 34.05 | 3.01 | 2.33 | 0 | 9.83 | 10.88 | 6.85 | 4.89 | |
| SUG-OccImage Backbone=ResNet-50, Params.=168.5M2026.01 | 14.92 | 35.87 | 10.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SymphoniesModality=C2026.06 | 14.9 | 41.9 | — | 56.4 | 27.6 | 15.3 | 1 | 21.6 | 28.7 | 20.4 | 2.5 | 2.8 | 13.9 | 25.7 | 6.6 | 30.9 | 3.5 | 2.2 | 0 | 8.4 | 9.6 | 5.8 | — | |
| Symphoniesevaluation_mode=single-frame2026.04 | 14.89 | — | — | 57.11 | 27.79 | 16.97 | 1.22 | 21.74 | 29.08 | 14.36 | 3.75 | 3.06 | 14.38 | 25.32 | 6.87 | 29.7 | 4.27 | 2.76 | 8 | 8.71 | 9.5 | 6.24 | 6.04 | |
| SymphoniesFramework=Two-Stage Methods2026.07 | 14.89 | 41.92 | 5.2 | 56.37 | 27.58 | 15.28 | 0.95 | 21.64 | 28.68 | 20.44 | 2.54 | 2.82 | 13.89 | 25.72 | 6.6 | 30.87 | 3.52 | 2.24 | 0 | 8.4 | 9.57 | 5.76 | — | |
| SGNevaluation_mode=single-frame2026.04 | 14.55 | — | — | 59.32 | 30.51 | 18.46 | 0.42 | 21.43 | 31.88 | 13.18 | 0.58 | 0.17 | 5.68 | 25.98 | 7.43 | 34.42 | 1.28 | 1.49 | 0 | 9.66 | 9.83 | 4.71 | 4.07 | |
| SparseOcc++ (Swin)Framework=Sparse Occupancy Methods, Backbone=Swin, Hardware=NVIDIA A100 GPU, Precision=FP32, Batch size=12026.07 | 13.97 | 37.32 | 20.1 | 59.08 | 29.43 | 22.06 | 0.5 | 15.29 | 25.68 | 18.45 | 1.34 | 3.07 | 12.72 | 19.34 | 3.28 | 31.82 | 4.17 | 4.63 | 0 | 7.48 | 4.29 | 2.75 | — | |
| LinkOccModality=C2026.06 | 13.9 | 42.1 | — | 55.1 | 27.9 | 22.6 | 0.5 | 20.3 | 28.8 | 18 | 2.3 | 1.5 | 10.1 | 24 | 6.9 | 32.9 | 3.8 | 5.1 | 0 | 6 | 7.3 | 4 | — | |
| ProtoOccVenue=AAAI'2025, Image Backbone=EfficientNet-B7, Params.=167.0M2026.01 | 13.89 | 36.67 | 6.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SAG-IRTModality=C2026.06 | 13.8 | 40.5 | — | 59.3 | 27.8 | 16.2 | 1.1 | 15.9 | 27.7 | 16.1 | 1.5 | 2.3 | 9.6 | 21 | 6.6 | 28.7 | 3.7 | 8.6 | 0 | 7.8 | 5.1 | 2.8 | — | |
| OccFormerModality=C2026.06 | 13.5 | 36.5 | — | 58.9 | 26.9 | 19.6 | 0.3 | 14.4 | 25.1 | 25.5 | 0.8 | 1.2 | 8.5 | 19.6 | 3.9 | 32.6 | 2.8 | 2.8 | 0 | 5.6 | 4.3 | 2.9 | — | |
| HASSCevaluation_mode=single-frame2026.04 | 13.48 | — | — | 57.05 | 28.25 | 15.9 | 1.04 | 19.05 | 27.23 | 9.91 | 0.92 | 0.86 | 5.61 | 25.48 | 6.15 | 32.94 | 2.8 | 4.71 | 0 | 6.58 | 7.68 | 4.05 | 3.97 | |
| OccFormerVenue=ICCV'2023, Image Backbone=EfficientNet-B7, Params.=203.4M2026.01 | 13.46 | 36.5 | 4.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OccFormerevaluation_mode=single-frame2026.04 | 13.46 | — | — | 58.85 | 26.88 | 19.61 | 0.31 | 14.4 | 25.09 | 25.53 | 0.81 | 1.19 | 8.52 | 19.63 | 3.93 | 32.62 | 2.78 | 2.82 | 0 | 5.61 | 4.26 | 2.86 | 4.82 | |
| OccFormerFramework=Dense Occupancy Methods2026.07 | 13.46 | 36.5 | 4.2 | 58.85 | 26.88 | 19.61 | 0.31 | 14.4 | 25.09 | 25.53 | 0.81 | 1.19 | 8.52 | 19.63 | 3.93 | 32.62 | 2.78 | 2.82 | 0 | 5.61 | 4.26 | 2.86 | — | |
| SparseOcc++Framework=Sparse Occupancy Methods, Hardware=NVIDIA A100 GPU, Precision=FP32, Batch size=12026.07 | 13.23 | 37.08 | 24.7 | 58.46 | 28.73 | 19.85 | 1.35 | 14.74 | 26.26 | 20.34 | 0.82 | 2.13 | 8.07 | 18.11 | 2.93 | 30.75 | 3.6 | 1.58 | 0 | 6.66 | 4.27 | 2.74 | — | |
| SparseOccVenue=CVPR'2024, Image Backbone=EfficientNet-B7, Params.=229.9M2026.01 | 13.12 | 36.67 | 6.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SparseOccFramework=Sparse Occupancy Methods, Hardware=NVIDIA A100 GPU, Precision=FP32, Batch size=12026.07 | 13.12 | 36.48 | 6.7 | 59.59 | 29.68 | 20.44 | 0.47 | 15.41 | 24.03 | 18.07 | 0.78 | 0.89 | 8.94 | 18.89 | 3.46 | 31.06 | 3.68 | 0.62 | 0 | 6.73 | 3.89 | 2.6 | — | |
| NDC-sceneevaluation_mode=single-frame2026.04 | 12.7 | — | — | 59.2 | 28.24 | 21.42 | 1.67 | 14.94 | 26.26 | 14.75 | 1.67 | 2.37 | 7.73 | 19.09 | 3.51 | 31.04 | 3.6 | 2.74 | 0 | 6.65 | 4.53 | 2.73 | 4.12 | |
| NDC-SceneFramework=Dense Occupancy Methods2026.07 | 12.7 | 37.24 | 4.5 | 59.2 | 28.24 | 21.42 | 1.67 | 14.94 | 26.26 | 14.75 | 1.67 | 2.37 | 7.73 | 19.09 | 3.51 | 31.04 | 3.6 | 2.74 | 0 | 6.65 | 4.53 | 2.73 | — | |
| VoxFormerModality=C2026.06 | 12.4 | 44 | — | 54.8 | 26.4 | 15.5 | 0.7 | 17.7 | 25.8 | 5.6 | 0.6 | 0.5 | 3.8 | 24.4 | 5.1 | 30 | 1.8 | 3.3 | 0 | 7.6 | 7.1 | 4.2 | — | |
| VoxFormer-SVenue=CVPR'2023, Image Backbone=ResNet-502026.01 | 12.35 | 44.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoxFormerevaluation_mode=single-frame2026.04 | 12.35 | — | — | 54.76 | 26.35 | 15.5 | 0.7 | 17.65 | 25.79 | 5.63 | 0.59 | 0.51 | 3.77 | 24.39 | 5.08 | 29.96 | 1.78 | 3.32 | 0 | 7.64 | 7.11 | 4.18 | 2.97 | |
| VoxFormerFramework=Two-Stage Methods2026.07 | 12.35 | 44.02 | 6 | 54.76 | 26.35 | 15.5 | 0.7 | 17.65 | 25.79 | 5.63 | 0.59 | 0.51 | 3.77 | 24.39 | 5.08 | 29.96 | 1.78 | 3.32 | 0 | 7.64 | 7.11 | 4.18 | — | |
| MonoSceneVenue=CVPR'2022, Image Backbone=EfficientNet-B72026.01 | 11.5 | 37.12 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MonoSceneevaluation_mode=single-frame2026.04 | 11.5 | — | — | 57.47 | 27.05 | 15.72 | 0.87 | 14.24 | 23.55 | 7.83 | 0.2 | 0.77 | 3.59 | 18.12 | 2.57 | 30.76 | 1.79 | 1.03 | 0 | 6.39 | 4.11 | 2.48 | 2.29 | |
| MonoSceneModality=C2026.06 | 11.5 | 37.1 | — | 57.5 | 27.1 | 15.7 | 0.9 | 14.2 | 23.6 | 7.8 | 0.2 | 0.8 | 3.6 | 18.1 | 2.6 | 30.8 | 1.8 | 1 | 0 | 6.4 | 4.1 | 2.5 | — | |
| TPVFormerModality=C2026.06 | 11.4 | 35.6 | — | 56.5 | 25.9 | 20.6 | 0.9 | 13.9 | 23.8 | 8.1 | 0.4 | 0.1 | 4.4 | 16.9 | 2.3 | 30.4 | 0.5 | 0.9 | 0 | 5.9 | 3.1 | 1.5 | — | |
| TPVFormerVenue=CVPR'2023, Image Backbone=ResNet-1012026.01 | 11.36 | 34.25 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TPVFormerevaluation_mode=single-frame2026.04 | 11.36 | — | — | 56.5 | 25.87 | 20.6 | 0.85 | 13.88 | 23.81 | 8.08 | 0.36 | 0.05 | 4.35 | 16.92 | 2.26 | 30.38 | 0.51 | 0.89 | 0 | 5.94 | 3.14 | 1.52 | 2 | |
| TPVFormerFramework=Dense Occupancy Methods2026.07 | 11.36 | 35.61 | — | 56.5 | 25.87 | 20.6 | 0.85 | 13.88 | 23.81 | 8.08 | 0.36 | 0.05 | 4.35 | 16.92 | 2.26 | 30.38 | 0.51 | 0.89 | 0 | 5.94 | 3.14 | 1.52 | — | |
| MonoSceneFramework=Dense Occupancy Methods2026.07 | 11.08 | 36.86 | 3.9 | 56.52 | 26.72 | 14.27 | 0.46 | 14.09 | 23.26 | 6.98 | 0.61 | 0.45 | 1.48 | 17.89 | 2.81 | 29.64 | 1.86 | 1.2 | 0 | 5.84 | 4.14 | 2.25 | — | |
| JS3C-Net*Venue=NeurIPS'20212026.01 | 10.31 | 38.98 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| JS3C-NetFramework=Dense Occupancy Methods, Reported by=[42]2026.07 | 10.31 | 38.98 | — | 50.49 | 23.74 | 11.94 | 0.07 | 15.03 | 24.65 | 4.41 | 0 | 0 | 6.15 | 18.11 | 4.33 | 26.86 | 0.67 | 0.27 | 0 | 3.94 | 3.77 | 1.45 | — | |
| AICNet*Venue=CVPR'20202026.01 | 8.31 | 29.59 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AICNetFramework=Dense Occupancy Methods, Reported by=[42]2026.07 | 8.31 | 29.59 | — | 43.55 | 20.55 | 11.97 | 0.07 | 12.94 | 14.71 | 4.53 | 0 | 0 | 0 | 15.37 | 2.9 | 28.71 | 0 | 0 | 0 | 2.52 | 0.06 | 0 | — | |
| 3DSketch*Venue=CVPR'20202026.01 | 7.5 | 33.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| 3DSketchFramework=Dense Occupancy Methods, Reported by=[42]2026.07 | 7.5 | 33.3 | — | 41.32 | 21.63 | 0 | 0 | 14.81 | 18.59 | 0 | 0 | 0 | 0 | 19.09 | 0 | 26.4 | 0 | 0 | 0 | 0.73 | 0 | 0 | — | |
| LMSCNet*Venue=3DV'20202026.01 | 6.7 | 28.61 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LMSCNetFramework=Dense Occupancy Methods, Reported by=[42]2026.07 | 6.7 | 28.61 | — | 40.68 | 18.22 | 4.38 | 0 | 10.31 | 18.33 | 0 | 0 | 0 | 0 | 13.66 | 0.02 | 20.54 | 0 | 0 | 0 | 1.21 | 0 | 0 | — |