3D Semantic Segmentation on SynLiDAR -> semanticPOSS (test)
50.8mIoUDGT-ST
Evaluation results
| Method | Links | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DGT-STMech.=self-training2024.03 | 50.8 | 55.1 | 70.7 | 46.1 | 74.2 | 30.1 | 36.3 | 44.1 | 81 | 4.3 | 62.8 | 10.3 | 78.5 | 67.2 | 12.5 | |
| PMANMech.=adversarial training2024.03 | 46.5 | 52.6 | 61.5 | 46.8 | 75.1 | 18.8 | 36.5 | 21.4 | 74.7 | 18.3 | 25.8 | 37.5 | 73.7 | 61.9 | 8.2 | |
| LaserMixMech.=self-training2024.03 | 45.5 | 58.4 | 61.3 | 47.7 | 69 | 21.9 | 39.5 | 30.9 | 61 | 16.1 | 36.5 | 7.1 | 79.5 | 62.6 | 7.2 | |
| CoSMixMech.=self-training2024.03 | 44.6 | 53.6 | 47.6 | 44.8 | 75.1 | 16.8 | 37.9 | 25.3 | 72.7 | 19.9 | 39.7 | 10.8 | 80 | 56.5 | 6.3 | |
| PCANMech.=adversarial training2024.03 | 44.4 | 48.6 | 62.1 | 37.5 | 74 | 23.9 | 31.4 | 22.2 | 76.9 | 6.5 | 41.9 | 11.9 | 79.1 | 61.2 | 6.1 | |
| ADVENTMech.=adversarial training2024.03 | 40.1 | 44.6 | 47.6 | 40.3 | 71.2 | 15.6 | 35.6 | 22 | 68.4 | 10.6 | 25.9 | 10.4 | 76.7 | 52.3 | 1.8 | |
| CLANMech.=adversarial training2024.03 | 39.5 | 43.9 | 46.6 | 41.3 | 71 | 15.1 | 34.3 | 20.4 | 69.6 | 9.5 | 23.2 | 12 | 75.1 | 51.3 | 1.2 | |
| MRNetMech.=adversarial training2024.03 | 39.4 | 43.5 | 47.2 | 39.1 | 70.4 | 15.5 | 32.8 | 22 | 66.1 | 13.2 | 24.2 | 11.2 | 76.8 | 50 | 1.1 | |
| AdaptSegNetMech.=adversarial training2024.03 | 39.3 | 43.9 | 48.2 | 39 | 69.6 | 15.5 | 33.6 | 21.3 | 64.3 | 12.7 | 25 | 11.6 | 76 | 49.9 | 1 | |
| Source onlyMech.=None2024.03 | 38.3 | 47.2 | 43.6 | 37.8 | 70.3 | 11.1 | 33.8 | 19.5 | 67.9 | 11.2 | 19.9 | 9.6 | 77.9 | 47.8 | 0 | |
| FADAMech.=adversarial training2024.03 | 37.6 | 39.6 | 41.2 | 38.8 | 69.2 | 16.3 | 32.1 | 18.1 | 67.9 | 11.5 | 22 | 13 | 71.4 | 47.9 | -0.7 | |
| PolarMixMech.=self-training2024.03 | 30.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | -8.3 |