Semantic Segmentation on Cityscapes fine-only (test)
81.8mIoUGALDNet
Evaluation results
| Method | Links | |
|---|---|---|
| GALDNetBackbone=ResNet101, Training Data=train-fine, Multi-scale crop test=true2019.09 | 81.8 | |
| DAnetBackbone=ResNet101, Training Data=train-fine, Multi-scale crop test=true2019.09 | 81.5 | |
| GloreBackbone=ResNet101, Training Data=train-fine, Multi-scale crop test=true2019.09 | 80.9 | |
| GALDNetBackbone=ResNet50, Training Data=train-fine, Multi-scale crop test=true2019.09 | 80.8 | |
| DenseASPPBackbone=DenseNet161, Training Data=train-fine, Multi-scale crop test=true2019.09 | 80.6 | |
| PSANetBackbone=ResNet101, Training Data=train-fine, Multi-scale crop test=true2019.09 | 80.1 | |
| GloreBackbone=ResNe50, Training Data=train-fine, Multi-scale crop test=true2019.09 | 79.5 | |
| DFNBackbone=ResNet101, Training Data=train-fine, Multi-scale crop test=true2019.09 | 79.3 | |
| AAFBackbone=ResNet101, Training Data=train-fine, Multi-scale crop test=true2019.09 | 79.1 | |
| BiSeNetBackbone=ResNet101, Training Data=train-fine, Multi-scale crop test=true2019.09 | 78.9 | |
| SACBackbone=ResNet101, Training Data=train-fine, Multi-scale crop test=true2019.09 | 78.1 |