Semantic Segmentation on Cityscapes fine+coarse (test)
83.3mIoUGALDNet(+Mapillary)
Evaluation results
| Method | Links | |
|---|---|---|
| GALDNet(+Mapillary)Backbone=ResNet101, Training Data=fine + coarse + Mapillary, Multi-scale crop test=true2019.09 | 83.3 | |
| GALDNet(+Mapillary)Backbone=ResNet101, Training Data=fine + coarse + Mapillary, Multi-scale crop test=true2019.09 | 83.3 | |
| GALDNetBackbone=ResNet101, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 82.9 | |
| GALDNetBackbone=ResNet101, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 82.9 | |
| DPCBackbone=Xception, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 82.6 | |
| DPCBackbone=Xception, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 82.6 | |
| Deeplabv3+Backbone=Xception, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 82.1 | |
| Auto-DeeplabBackbone=Auto-Deeplab Architecture, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 82.1 | |
| Deeplabv3+Backbone=Xception, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 82.1 | |
| Auto-DeeplabBackbone=AutoML Search Space, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 82.1 | |
| PSPBackbone=ResNet101, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 81.2 | |
| PSPBackbone=ResNet101, Training Data=fine + coarse, Multi-scale crop test=true2019.09 | 81.2 |