Semantic Segmentation on Cityscapes (val) (s.s. mIoU and m.s. mIoU)
86.9mIoU (s.s.)HRNetV2-OCR+PSA
Evaluation results
| Method | Links | ||
|---|---|---|---|
| HRNetV2-OCR+PSABackbone=HRNetV2-W48, Crop Size=1024x2048, Extra Data=false2022.11 | 86.9 | — | |
| HRNetV2-OCRBackbone=HRNetV2-W48, Crop Size=1024x2048, Extra Data=true2022.11 | 86.3 | — | |
| Panoptic-DeepLabBackbone=SWideRNet, Crop Size=1025x2049, Extra Data=true2022.11 | 85.3 | — | |
| Mask2FormerBackbone=ViT-Adapter-L, Number of parameters=571M, Crop Size=896x896, Extra Data=true2022.11 | 84.9 | 85.8 | |
| Axial-DeepLab-XLBackbone=Axial ResNet-XL, Number of parameters=173M, Crop Size=1025x2049, Extra Data=true2022.11 | 84.6 | — | |
| Panoptic-DeepLabBackbone=SWideRNet, Crop Size=1025x2049, Extra Data=true2022.11 | 84.6 | 85.3 | |
| Axial-DeepLab-XLBackbone=Axial ResNet-XL, Number of parameters=123M, Crop Size=1025x2049, Extra Data=true2022.11 | 84.2 | — | |
| Mask2FormerBackbone=SeMask Swin-L, Number of parameters=223M, Crop Size=512x1024, Extra Data=false2022.11 | 84 | 85 | |
| OneFormerBackbone=ConvNeXt-XL, Number of parameters=372M, Crop Size=512x1024, Extra Data=false2022.11 | 83.6 | 84.6 | |
| kMaX-DeepLabBackbone=ConvNeXt-L, Number of parameters=232M, Crop Size=1025x2049, Extra Data=false2022.11 | 83.5 | — | |
| Mask2Former-SemanticBackbone=Swin-L, Number of parameters=215M, Crop Size=512x1024, Extra Data=false2022.11 | 83.3 | 84.3 | |
| OneFormerBackbone=DINAT-L, Number of parameters=223M, Crop Size=512x1024, Extra Data=false2022.11 | 83.1 | 84 | |
| OneFormerBackbone=Swin-L, Number of parameters=219M, Crop Size=512x1024, Extra Data=false2022.11 | 83 | 84.4 | |
| OneFormerBackbone=ConvNeXt-L, Number of parameters=220M, Crop Size=512x1024, Extra Data=false2022.11 | 83 | 84 | |
| Mask2Former-PanopticBackbone=Swin-L, Number of parameters=216M, Crop Size=512x1024, Extra Data=false2022.11 | 82.9 | — | |
| Panoptic-DeepLabBackbone=SWideRNet, Crop Size=1025x2049, Extra Data=false2022.11 | 82.2 | 82.9 | |
| EfficientPSBackbone=EfficientNet, Crop Size=1025x2049, Extra Data=true2022.11 | 82.1 | — | |
| Axial-DeepLab-XLBackbone=Axial ResNet-XL, Number of parameters=173M, Crop Size=1025x2049, Extra Data=false2022.11 | 80.6 | 81.1 |