LiDAR-based Panoptic Segmentation on nuScenes (val)
74.7PQPanoptic-PHNet
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Panoptic-PHNetTraining Type=Single-Dataset Training2024.05 | 74.7 | 77.7 | 84.2 | 88.2 | — | — | — | — | — | — | 79.7 | |
| Panoptic-PHNet2022.05 | 74.7 | 77.7 | 84.2 | 88.2 | 74 | 82.5 | 89 | 75.9 | 86.9 | 86.8 | 79.7 | |
| M3NetData alignment=true, Feature alignment=true, Label alignment=true2024.05 | 71.7 | 74.01 | 82.2 | 86.47 | — | — | — | — | — | — | 80.9 | |
| M3NetData alignment=true, Feature alignment=false, Label alignment=true2024.05 | 71.53 | 73.91 | 81.8 | 86.92 | — | — | — | — | — | — | 80.45 | |
| M3NetData alignment=true, Feature alignment=true, Label alignment=false2024.05 | 71.47 | 73.86 | 81.53 | 86.71 | — | — | — | — | — | — | 80.26 | |
| M3NetData alignment=true, Feature alignment=false, Label alignment=false2024.05 | 68.49 | 71.11 | 79.13 | 85.49 | — | — | — | — | — | — | 79.13 | |
| DSNetTraining Type=Single-Dataset Training2024.05 | 64.7 | 67.6 | 76.1 | 83.5 | — | — | — | — | — | — | 76.3 | |
| Panoptic-PolarNetTraining Type=Single-Dataset Training2024.05 | 63.4 | 67.2 | 75.3 | 83.9 | — | — | — | — | — | — | 66.9 | |
| Panoptic-PolarNet2022.05 | 63.4 | 67.2 | 75.3 | 83.9 | 59.2 | 70.3 | 84.1 | 70.4 | 83.5 | 83.6 | 66.9 | |
| EfficientLPS2022.05 | 62 | 65.6 | 73.9 | 83.4 | 56.8 | 68 | 83.2 | 70.6 | 83.6 | 83.8 | 65.6 | |
| EfficientLPSTraining Type=Single-Dataset Training2024.05 | 59.2 | 62.8 | 82.9 | 70.7 | — | — | — | — | — | — | 69.4 | |
| BaselineTraining Type=Naïve Joint Training2024.05 | 56.67 | 60.61 | 66.75 | 83.49 | — | — | — | — | — | — | 69.65 | |
| Panoptic-TrackNetTraining Type=Single-Dataset Training2024.05 | 51.4 | 56.3 | 63.3 | 80.2 | — | — | — | — | — | — | 58 | |
| PanopticTrackNet2022.05 | 51.4 | 56.2 | 63.3 | 80.2 | 45.8 | 55.9 | 81.4 | 60.4 | 75.5 | 78.3 | 58 | |
| DS-Net2020.11 | 42.5 | 51 | 50.3 | 83.6 | 32.5 | 38.3 | 83.1 | 59.2 | 70.3 | 84.4 | 70.7 | |
| Cylinder3D + SECONDSemantic Segmentation Model=Cylinder3D, Instance Segmentation Model=SECOND2020.11 | 40.1 | 48.4 | 47.3 | 84.2 | 29 | 33.6 | 84.4 | 58.5 | 70.1 | 83.7 | 58.5 | |
| Cylinder3D + PointPillarsSemantic Segmentation Model=Cylinder3D, Instance Segmentation Model=PointPillars2020.11 | 36 | 44.5 | 43 | 83.3 | 23.3 | 27 | 83.7 | 57.2 | 69.6 | 82.7 | 52.3 |