Semantic Occupancy Prediction on EmbodiedScan (test)
20.79mIoUMulti-Modality
Evaluation results
| Method | Links | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Multi-ModalityInput=RGB-D, Setting=Continuous2023.12 | 20.79 | 73.5 | 63.64 | 62.3 | 54.6 | 19.96 | 48.99 | 61.1 | 69.76 | 39.86 | 34.62 | 54.83 | 54.45 | 48.9 | 41.22 | 7.97 | 63.52 | |
| Multi-ModalityInput=RGB-D, Setting=Multi-view2023.12 | 19.97 | 71.21 | 64.92 | 55 | 52.04 | 27.35 | 33.97 | 47.93 | 46.26 | 31.87 | 27.98 | 46.58 | 46.56 | 24.05 | 39.01 | 24.4 | 67.79 | |
| Depth-OnlyInput=Depth, Setting=Multi-view2023.12 | 15.56 | 69.92 | 60.52 | 51.74 | 49.44 | 23.08 | 24.33 | 45.77 | 43.52 | 29.74 | 23.02 | 39.04 | 41.22 | 17.42 | 19.58 | 25.79 | 60.45 | |
| Depth-OnlyInput=Depth, Setting=Continuous2023.12 | 14.44 | 73.91 | 66.22 | 56.13 | 49.96 | 15.7 | 24.37 | 56.84 | 55.35 | 30.55 | 26.66 | 42.81 | 30.81 | 33.01 | 21.71 | 6.21 | 45.35 | |
| Camera-OnlyInput=RGB, Setting=Multi-view2023.12 | 10.48 | 40.45 | 41.25 | 27.19 | 26.16 | 15.5 | 20.3 | 30.82 | 26.7 | 15.01 | 14.33 | 29.17 | 23.3 | 16.99 | 15.98 | 6.17 | 42.57 | |
| Camera-OnlyInput=RGB, Setting=Continuous2023.12 | 10.43 | 39.09 | 34.1 | 30.24 | 26.46 | 9.49 | 25.53 | 41.6 | 35.19 | 16.22 | 20.45 | 24.46 | 19.8 | 26.01 | 17.26 | 1.83 | 29.78 | |
| SurroundOccInput=RGB, Setting=Multi-view2023.12 | 9.1 | 38.54 | 46.17 | 23.55 | 23.04 | 13.6 | 19.15 | 27.79 | 22.88 | 13.11 | 13.72 | 24.32 | 18.89 | 13.58 | 14.77 | 7.83 | 34.71 | |
| OccNetInput=RGB, Setting=Multi-view2023.12 | 8.07 | 37.15 | 46.9 | 25.63 | 20.94 | 13.17 | 18.4 | 26.81 | 22.86 | 13.59 | 13.49 | 26.75 | 22.92 | 17.15 | 17.07 | 4.77 | 33.6 |