3D Semantic Occupancy Prediction on nuScenes 1.0 (val)
21.39IoU (barrier)GaussianFormer-2
Evaluation results
| Method | Links | ||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GaussianFormer-2Channel dimension (Ch.)=1922024.12 | 21.39 | 31.74 | 20.82 | 13.44 | 28.49 | 30.82 | 10.92 | 15.84 | 13.55 | 10.53 | 14.04 | 22.92 | 40.61 | 24.36 | 26.08 | 24.27 | 13.83 | 21.98 | — | — | |
| GaussianWorld2024.12 | 21.38 | 33.4 | 22.13 | 1,412 | 27.71 | 31.84 | 13.66 | 1,743 | 13.66 | 1,146 | 15.09 | 23.94 | 42.98 | 24.86 | 28.84 | 26.74 | 15.69 | 24.74 | — | — | |
| GaussianFormer-TMode=temporal fusion2024.12 | 20.82 | 31.34 | 20.42 | 1,207 | 26.89 | 30.94 | 10.52 | 1,648 | 13.15 | 1,046 | 12.9 | 21.79 | 41.13 | 24.22 | 26.29 | 24.89 | 12.8 | 21.45 | — | — | |
| VG3SVenue=Ours2026.03 | 20.78 | — | — | 12.4 | 28.09 | 29.53 | 11.93 | 15.59 | 12.38 | 10.61 | 14.65 | 21.74 | 42.42 | 26.39 | 28.06 | 26.58 | 17.46 | 25.76 | 34.41 | 21.52 | |
| SurroundOcc2024.12 | 20.59 | 31.49 | 20.3 | 11.68 | 28.06 | 30.86 | 10.7 | 15.14 | 14.09 | 12.06 | 14.38 | 22.26 | 37.29 | 23.7 | 24.49 | 22.77 | 14.89 | 21.86 | — | — | |
| SurroundOcc2024.12 | 20.59 | 31.49 | 20.3 | 1,168 | 28.06 | 30.86 | 10.7 | 1,514 | 14.09 | 1,206 | 14.38 | 22.26 | 37.29 | 23.7 | 24.49 | 22.77 | 14.89 | 21.86 | — | — | |
| SurroundOccVenue=ICCV 20232026.03 | 20.59 | — | — | 11.68 | 28.06 | 30.86 | 10.7 | 15.14 | 14.09 | 12.06 | 14.38 | 22.26 | 37.29 | 23.7 | 24.49 | 22.77 | 14.89 | 21.86 | 31.49 | 20.3 | |
| GaussianFormer-2Channel dimension (Ch.)=1282024.12 | 20.15 | 30.56 | 20.02 | 12.99 | 27.61 | 30.23 | 11.19 | 15.31 | 12.64 | 9.63 | 13.31 | 22.26 | 39.68 | 23.47 | 25.62 | 23.2 | 12.25 | 20.73 | — | — | |
| GaussianFormer-2Venue=CVPR 2025, feature dimension=128-channel2026.03 | 20.15 | — | — | 12.99 | 27.61 | 30.23 | 11.19 | 15.31 | 12.64 | 9.63 | 13.31 | 22.26 | 39.68 | 23.47 | 25.62 | 23.2 | 12.25 | 20.73 | 30.56 | 20.02 | |
| GaussianFormer2024.12 | 19.52 | 29.83 | 19.1 | 11.26 | 26.11 | 29.78 | 10.47 | 13.83 | 12.58 | 8.67 | 12.74 | 21.57 | 39.63 | 23.28 | 24.46 | 22.99 | 9.59 | 19.12 | — | — | |
| GaussianFormer2024.12 | 19.52 | 29.83 | 19.1 | 1,126 | 26.11 | 29.78 | 10.47 | 1,383 | 12.58 | 867 | 12.74 | 21.57 | 39.63 | 23.28 | 24.46 | 22.99 | 9.59 | 19.12 | — | — | |
| GaussianFormerVenue=ECCV 20242026.03 | 19.52 | — | — | 11.26 | 26.11 | 29.78 | 10.47 | 13.83 | 12.58 | 8.67 | 12.74 | 21.57 | 39.63 | 23.28 | 24.46 | 22.99 | 9.59 | 19.12 | 29.83 | 19.1 | |
| GaussianFormer-BMode=single-frame2024.12 | 19.36 | 30.68 | 19.73 | 1,319 | 26.9 | 29.79 | 10.2 | 1,517 | 12.55 | 929 | 12.96 | 21.45 | 39.55 | 23.03 | 25.07 | 23.65 | 12.35 | 21.18 | — | — | |
| OccFormer2024.12 | 18.65 | 31.39 | 19.03 | 10.41 | 23.92 | 30.29 | 10.31 | 14.19 | 13.59 | 10.13 | 12.49 | 20.77 | 38.78 | 19.79 | 24.19 | 22.21 | 13.48 | 21.35 | — | — | |
| OccFormer2024.12 | 18.65 | 31.39 | 19.03 | 1,041 | 23.92 | 30.29 | 10.31 | 1,419 | 13.59 | 1,013 | 12.49 | 20.77 | 38.78 | 19.79 | 24.19 | 22.21 | 13.48 | 21.35 | — | — | |
| OccFormerVenue=ICCV 20232026.03 | 18.65 | — | — | 10.41 | 23.92 | 30.29 | 10.31 | 14.19 | 13.59 | 10.13 | 12.49 | 20.77 | 38.78 | 19.79 | 24.19 | 22.21 | 13.48 | 21.35 | 31.39 | 19.03 | |
| TPVFormersupervision=original LiDAR segmentation labels2024.12 | 16.14 | 11.51 | 11.66 | 7.17 | 22.63 | 17.13 | 8.83 | 11.39 | 10.46 | 8.23 | 9.43 | 17.02 | 8.07 | 13.64 | 13.85 | 10.34 | 4.9 | 7.37 | — | — | |
| TPVFormersupervision=sparse LiDAR segmentation labels2024.12 | 16.14 | 11.51 | 11.66 | 717 | 22.63 | 17.13 | 8.83 | 1,139 | 10.46 | 823 | 9.43 | 17.02 | 8.07 | 13.64 | 13.85 | 10.34 | 4.9 | 7.37 | — | — | |
| TPVFormerVenue=CVPR 20232026.03 | 16.14 | — | — | 7.17 | 22.63 | 17.13 | 8.83 | 11.39 | 10.46 | 8.23 | 9.43 | 17.02 | 8.07 | 13.64 | 13.85 | 10.34 | 4.9 | 7.37 | 11.51 | 11.66 | |
| TPVFormersupervision=dense occupancy annotations2024.12 | 15.96 | 30.86 | 17.1 | 5.31 | 23.86 | 27.32 | 9.79 | 8.74 | 7.09 | 5.2 | 10.97 | 19.22 | 38.87 | 21.25 | 24.26 | 23.15 | 11.73 | 20.81 | — | — | |
| TPVFormer*supervision=dense occupancy labels2024.12 | 15.96 | 30.86 | 17.1 | 531 | 23.86 | 27.32 | 9.79 | 874 | 7.09 | 520 | 10.97 | 19.22 | 38.87 | 21.25 | 24.26 | 23.15 | 11.73 | 20.81 | — | — | |
| TPVFormerVenue=CVPR 2023, supervision using dense occupancy annotations=true2026.03 | 15.96 | — | — | 5.31 | 23.86 | 27.32 | 9.79 | 8.74 | 7.09 | 5.2 | 10.97 | 19.22 | 38.87 | 21.25 | 24.26 | 23.15 | 11.73 | 20.81 | 30.86 | 17.1 | |
| BEVFormer2024.12 | 14.22 | 30.5 | 16.75 | 6.58 | 23.46 | 28.28 | 8.66 | 10.77 | 6.64 | 4.05 | 11.2 | 17.78 | 37.28 | 18 | 22.88 | 22.17 | 13.8 | 22.21 | — | — | |
| BEVFormer2024.12 | 14.22 | 30.5 | 16.75 | 658 | 23.46 | 28.28 | 8.66 | 1,077 | 6.64 | 405 | 11.2 | 17.78 | 37.28 | 18 | 22.88 | 22.17 | 13.8 | 22.21 | — | — | |
| BEVFormerVenue=ECCV 20222026.03 | 14.22 | — | — | 6.58 | 23.46 | 28.28 | 8.66 | 10.77 | 6.64 | 4.05 | 11.2 | 17.78 | 37.28 | 18 | 22.88 | 22.17 | 13.8 | 22.21 | 30.5 | 16.75 | |
| Atlas2024.12 | 10.64 | 28.66 | 15 | 5.68 | 19.66 | 24.94 | 8.9 | 8.84 | 6.47 | 3.28 | 10.42 | 16.21 | 34.86 | 15.46 | 21.89 | 20.95 | 11.21 | 20.54 | — | — | |
| Atlas2024.12 | 10.64 | 28.66 | 15 | 568 | 19.66 | 24.94 | 8.9 | 884 | 6.47 | 328 | 10.42 | 16.21 | 34.86 | 15.46 | 21.89 | 20.95 | 11.21 | 20.54 | — | — | |
| AtlasVenue=ECCV 20202026.03 | 10.64 | — | — | 5.68 | 19.66 | 24.94 | 8.9 | 8.84 | 6.47 | 3.28 | 10.42 | 16.21 | 34.86 | 15.46 | 21.89 | 20.95 | 11.21 | 20.54 | 28.66 | 15 | |
| MonoScene2024.12 | 4.03 | 23.96 | 7.31 | 0.35 | 8 | 8.04 | 2.9 | 0.28 | 1.16 | 0.67 | 4.01 | 4.35 | 27.72 | 5.2 | 15.13 | 11.29 | 9.03 | 14.86 | — | — | |
| MonoScene2024.12 | 4.03 | 23.96 | 7.31 | 35 | 8 | 8.04 | 2.9 | 28 | 1.16 | 67 | 4.01 | 4.35 | 27.72 | 5.2 | 15.13 | 11.29 | 9.03 | 14.86 | — | — | |
| MonoSceneVenue=CVPR 20222026.03 | 4.03 | — | — | 0.35 | 8 | 8.04 | 2.9 | 0.28 | 1.16 | 0.67 | 4.01 | 4.35 | 27.72 | 5.2 | 15.13 | 11.29 | 9.03 | 14.86 | 23.96 | 7.31 |