Semantic Segmentation on Cross-City Rome
84.9Road mIoUFADA
Evaluation results
| Method | Links | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| FADABackbone=DeepLab-v22020.07 | 84.9 | 35.8 | 88.3 | 2,050 | 40.1 | 85.9 | 92.8 | 56.2 | 23.2 | 83.6 | 31.8 | 53.2 | 1,460 | 54.7 | |
| BaselineBackbone=ResNet-101, Adaptation Level=None2018.02 | 83.9 | 34.3 | 87.7 | 13 | 41.9 | 84.6 | 92.5 | 37.7 | 22.4 | 80.8 | 38.1 | 39.1 | 5.3 | 50.9 | |
| AdaptSegNetBackbone=ResNet-101, Adaptation Level=output space2018.02 | 83.9 | 34.2 | 88.3 | 18.8 | 40.2 | 86.2 | 93.1 | 47.8 | 21.7 | 80.9 | 47.8 | 48.3 | 8.6 | 53.8 | |
| Source DeepLab-v2Backbone=DeepLab-v2, Adaptation=None2020.07 | 83.9 | 34.3 | 87.7 | 1,300 | 41.9 | 84.6 | 92.5 | 37.7 | 22.4 | 80.8 | 38.1 | 39.1 | 530 | 50.9 | |
| AdaptSegNetBackbone=DeepLab-v22020.07 | 83.9 | 34.2 | 88.3 | 1,880 | 40.2 | 86.2 | 93.1 | 47.8 | 21.7 | 80.9 | 47.8 | 48.3 | 860 | 53.8 | |
| Cross-City2018.02 | 79.5 | 29.3 | 84.5 | 0 | 22.2 | 80.6 | 82.8 | 29.5 | 13 | 71.7 | 37.5 | 25.9 | 1 | 42.9 | |
| Cross-CityBackbone=Dilation-Frontend2020.07 | 79.5 | 29.3 | 84.5 | 0 | 22.2 | 80.6 | 82.8 | 29.5 | 13 | 71.7 | 37.5 | 25.9 | 100 | 42.9 | |
| AdaptSegNetBackbone=ResNet-101, Adaptation Level=feature space2018.02 | 78.8 | 28.6 | 85.5 | 16.6 | 40.1 | 85.3 | 79.6 | 42.4 | 20.7 | 79.6 | 58.8 | 45.5 | 6.1 | 51.4 | |
| Source Dilation-FrontendBackbone=Dilation-Frontend, Adaptation=None2020.07 | 77.7 | 21.9 | 81.1 | 10 | 10.7 | 78.9 | 88.1 | 21.6 | 10 | 67.2 | 30.4 | 6.1 | 60 | 38.2 |