3D Semantic Occupancy Prediction on OpenOccupancy
24.5mIoUALOcc-2D-GF
Evaluation results
| Method | Links | |||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ALOcc-2D-GFInput=C&D, Backbone=ResNet-50, Input Img. Size=900 x 1600, FPS=0.8, Memory=138912025.04 | 24.5 | 21.6 | 34.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OccLoffInput=C&L, Backbone=ResNet-101, Input Img. Size=900 x 1600, FPS=3.5, Memory=80892025.04 | 22.9 | 21.1 | 31.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ALOcc-2D*Input=C&D, Backbone=ResNet-50, Input Img. Size=900 x 1600, FPS=0.8, Memory=138392025.04 | 22.4 | 20.3 | 30.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| M-OccGenInput=C&L, Backbone=ResNet-50, Input Img. Size=900 x 1600, FPS=2.32025.04 | 22 | 20 | 30.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Co-OccInput=C&L, Backbone=ResNet-101, Input Img. Size=900 x 16002025.04 | 21.9 | 19.5 | 30.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| M-CONetInput=C&L, Backbone=ResNet-50, Input Img. Size=900 x 1600, FPS=2.92025.04 | 20.1 | 18 | 29.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ALOcc-2D-GFInput=C, Backbone=ResNet-50, Input Img. Size=900 x 1600, FPS=0.8, Memory=138572025.04 | 17.9 | 13.7 | 28.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| L-OccGenInput=L, FPS=3.72025.04 | 16.8 | 11.8 | 31.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| L-OccGenNumber of parameters=62M, Training status=end-to-end2026.06 | 16.8 | — | 31.6 | 18.8 | 5.1 | 14.8 | 19.6 | 7 | 7.7 | 11.5 | 6.7 | 13.9 | 14.6 | 36.4 | 22.1 | 22.8 | 22.3 | 20.6 | 24.5 | |
| L-CONetInput=L, FPS=42025.04 | 15.8 | 10.8 | 30.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| L-CONetNumber of parameters=63M, Training status=end-to-end2026.06 | 15.8 | — | 30.9 | 17.5 | 5.2 | 13.3 | 18.1 | 7.8 | 5.4 | 9.6 | 5.6 | 13.2 | 13.6 | 34.9 | 21.5 | 22.4 | 21.7 | 19.2 | 23.5 | |
| ALOcc-2D*Input=C, Backbone=ResNet-50, Input Img. Size=900 x 1600, FPS=0.8, Memory=138052025.04 | 15.7 | 11.5 | 25.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LightOccModality=Camera2024.12 | 15.6 | — | 25.4 | 16.2 | 9.7 | 15.1 | 18.5 | 9.2 | 12.3 | 12.1 | 8.3 | 7.8 | 14.2 | 33.1 | 22.5 | 22.7 | 20.9 | 11.4 | 15.9 | |
| C-OccGenInput=C, Backbone=ResNet-50, Input Img. Size=900 x 1600, FPS=32025.04 | 14.5 | 11.7 | 23.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OccGenModality=Camera2024.12 | 14.5 | — | 23.4 | 15.5 | 9.1 | 15.3 | 19.2 | 7.3 | 11.3 | 11.8 | 8.9 | 5.9 | 13.7 | 34.8 | 22 | 21.8 | 19.5 | 6 | 9.9 | |
| OccMambaModality=Camera2024.12 | 13.2 | — | 21.7 | 13.2 | 7.4 | 15.3 | 17.6 | 6.1 | 10.3 | 10.3 | 6.5 | 6.2 | 13.3 | 32.9 | 20.5 | 19.8 | 17.4 | 4.8 | 8.7 | |
| HilDANumber of parameters=39.5M (39.1M backbone + 0.4M head), Backbone training status=frozen, Decoder head=lightweight (0.4M)2026.06 | 13 | — | 30 | 11.6 | 1.5 | 7.9 | 19.6 | 3.5 | 4.6 | 22.6 | 5.8 | 5.3 | 11.2 | 28.4 | 8.9 | 13.2 | 14.5 | 23.3 | 26.2 | |
| C-CONetInput=C, Backbone=ResNet-50, Input Img. Size=900 x 1600, FPS=3.52025.04 | 12.8 | 10.6 | 20.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CONetModality=Camera2024.12 | 12.8 | — | 20.1 | 13.2 | 8.1 | 15.4 | 17.2 | 6.3 | 11.2 | 10 | 8.3 | 4.7 | 12.1 | 31.4 | 18.8 | 18.7 | 16.3 | 4.8 | 8.2 | |
| L-OpenOccupancyNumber of parameters=63M, Training status=end-to-end2026.06 | 11.7 | — | 30.8 | 12.2 | 4.2 | 11 | 12.2 | 8.3 | 4.4 | 8.7 | 4 | 8.4 | 10.3 | 23.5 | 16 | 14.9 | 15.7 | 15 | 17.9 | |
| 3DSketch+Input=C&D, Backbone=ResNet-50, Input Img. Size=900 x 16002025.04 | 10.7 | 7.4 | 25.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AICNet+Input=C&D, Backbone=ResNet-50, Input Img. Size=900 x 16002025.04 | 10.6 | 7.4 | 23.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OpenOccupancyModality=Camera2024.12 | 10.3 | — | 19.3 | 9.9 | 6.8 | 11.2 | 11.5 | 6.3 | 8.4 | 8.6 | 4.3 | 4.2 | 9.9 | 22 | 15.8 | 14.1 | 13.5 | 7.3 | 10.2 | |
| TPVFormerInput=C, Backbone=ResNet-50, Input Img. Size=900 x 16002025.04 | 7.8 | 11.7 | 15.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TPVFormerModality=Camera2024.12 | 7.8 | — | 15.3 | 9.3 | 4.1 | 11.3 | 10.1 | 5.2 | 4.3 | 5.9 | 5.3 | 6.8 | 6.5 | 13.6 | 9 | 8.3 | 8 | 9.2 | 8.2 | |
| MonoSceneModality=Camera2024.12 | 6.9 | — | 18.4 | 7.1 | 3.9 | 9.3 | 7.2 | 5.6 | 3 | 5.9 | 4.4 | 4.9 | 4.2 | 14.9 | 6.3 | 7.9 | 7.4 | 10 | 7.6 |