Semantic Segmentation on C-Driving, Cityscapes, KITTI, and WildDash Open (test)
46.9mIoU (C-Driving)Object Style Compensation
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Object Style CompensationType=OCDA, Train: Source=GTA5, Train: Target=C-Driving2023.09 | 46.9 | 43.6 | 46.5 | 40.1 | 44.3 | |
| ML-BPMType=OCDA, Train: Source=GTA5, Train: Target=C-Driving2023.09 | 42.5 | 41.7 | 44.3 | 34.6 | 40.8 | |
| ASTType=OCDA, Train: Source=GTA5, Train: Target=C-Driving2023.09 | 40.7 | 40.3 | 41.9 | 32.2 | 38.8 | |
| DACSType=UDA, Train: Source=GTA5, Train: Target=C-Driving2023.09 | 39.7 | 37 | 40.2 | 30.7 | 36.9 | |
| DHAType=OCDA, Train: Source=GTA5, Train: Target=C-Driving2023.09 | 39.4 | 38.8 | 40.1 | 30.9 | 37.5 | |
| CSFUType=OCDA, Train: Source=GTA5, Train: Target=C-Driving2023.09 | 38.9 | 38.6 | 37.9 | 29.1 | 36.1 | |
| RobustNetType=DG, Train: Source=GTA5, Train: Target=C-Driving2023.09 | 38.1 | 38.3 | 40.5 | 30.8 | 37 |