Semantic Segmentation on BDD100K (val)
67.59mIoUKnet-SWIN-L
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Knet-SWIN-LBackbone=Knet-SWIN-L, Calibration=Uncalibrated2022.12 | 67.59 | 0.065 | |
| Knet-SWIN-L (Selective Scaling)Backbone=Knet-SWIN-L, Calibration=Selective Scaling2022.12 | 67.59 | 0.04 | |
| ConvNeXt-LBackbone=ConvNeXt-L, Calibration=Uncalibrated2022.12 | 67.26 | 0.064 | |
| ConvNeXt-L (Selective Scaling)Backbone=ConvNeXt-L, Calibration=Selective Scaling2022.12 | 67.26 | 0.038 | |
| SegFormer-B5Backbone=SegFormer-B5, Calibration=Uncalibrated2022.12 | 65.08 | 0.064 | |
| SegFormer-B5 (Selective Scaling)Backbone=SegFormer-B5, Calibration=Selective Scaling2022.12 | 65.08 | 0.04 | |
| Knet-DeepLabBackbone=Knet-DeepLab, Calibration=Uncalibrated2022.12 | 62.89 | 0.06 | |
| Knet-DeepLab (Selective Scaling)Backbone=Knet-DeepLab, Calibration=Selective Scaling2022.12 | 62.89 | 0.035 | |
| Segmenter-LBackbone=Segmenter-L, Calibration=Uncalibrated2022.12 | 61.33 | 0.055 | |
| Segmenter-L (Selective Scaling)Backbone=Segmenter-L, Calibration=Selective Scaling2022.12 | 61.33 | 0.031 | |
| ReinEncoder Parameters=307M, DG Category=Vision-Language DG2023.12 | 60.5 | — | |
| VLTSegEncoder Parameters=304M, DG Category=Vision-Language DG2023.12 | 58.3 | — | |
| CLOUDSEncoder Parameters=198M, DG Category=Vision-Language DG2023.12 | 57.4 | — | |
| DIDEXEncoder=Transformer, No real data usage=true, UDA component=MIC [17]2023.12 | 54.3 | — | |
| DIDEXEncoder Parameters=81.4M, DG Category=Vision-Language DG2023.12 | 54.3 | — | |
| SANSAWBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=Cityscapes2024.03 | 52.95 | — | |
| DGinStyleEncoder Parameters=81.4M, DG Category=Vision-Language DG2023.12 | 52.3 | — | |
| BlindNetBackbone=ResNet-50, External Dataset=false, External Module=true, Source Domain=Cityscapes2024.03 | 51.84 | — | |
| DIRLBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=Cityscapes2024.03 | 51.8 | — | |
| SiamDoGeBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=Cityscapes2024.03 | 51.53 | — | |
| DPCLBackbone=ResNet-50, External Dataset=false, External Module=true, Source Domain=Cityscapes2024.03 | 50.97 | — | |
| WildNetBackbone=ResNet-50, External Dataset=true, External Module=false, Source Domain=Cityscapes2024.03 | 50.94 | — | |
| RobustNetBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=Cityscapes2024.03 | 50.73 | — | |
| CMFormerSource Dataset=GTA52023.07 | 49.91 | — | |
| CMFormerEncoder=Transformer, No real data usage=true2023.12 | 49.9 | — | |
| CMFormerEncoder Parameters=197M, DG Category=Vision DG2023.12 | 49.9 | — | |
| IBAFormerEncoder Parameters=81.4M, DG Category=Vision DG2023.12 | 49.8 | — | |
| HRDAEncoder=Transformer, No real data usage=true2023.12 | 49.1 | — | |
| HRDAEncoder Parameters=81.4M, DG Category=Vision DG2023.12 | 49.1 | — | |
| IBN-NetBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=Cityscapes2024.03 | 48.56 | — | |
| SHADEEncoder Parameters=81.4M, DG Category=Vision DG2023.12 | 48.2 | — | |
| ReVTEncoder=Transformer, No real data usage=true2023.12 | 48 | — | |
| ReVTEncoder Parameters=81.4M, DG Category=Vision DG2023.12 | 48 | — | |
| DAFormerEncoder=Transformer, No real data usage=true2023.12 | 47.9 | — | |
| BaselineEncoder=Transformer, No real data usage=true2023.12 | 45.6 | — | |
| BaselineDG Category=Vision DG2023.12 | 45.6 | — | |
| RICAEncoder=ResNet, No real data usage=true2023.12 | 45.2 | — | |
| BaselineBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=Cityscapes2024.03 | 44.96 | — | |
| SHADEEncoder=ResNet, No real data usage=true2023.12 | 43.7 | — | |
| SHADE*Venue=IJCV 24, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=true2025.03 | 43.66 | — | |
| DRSFVenue=-, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=false2025.03 | 43.5 | — | |
| DGInStyleVenue=ECCV 24, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=false2025.03 | 42.81 | — | |
| WildNet*Venue=CVPR 22, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=true2025.03 | 41.73 | — | |
| WildNetEncoder=ResNet, No real data usage=false2023.12 | 41.7 | — | |
| BlindNetBackbone=ResNet-50, External Dataset=false, External Module=true, Source Domain=GTAV2024.03 | 41.32 | — | |
| FSDREncoder=ResNet, No real data usage=false2023.12 | 41.2 | — | |
| SAN+SAWEncoder=ResNet, No real data usage=true2023.12 | 41.2 | — | |
| FSDRVenue=CVPR 21, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=false2025.03 | 41.2 | — | |
| SAN-SAWVenue=CVPR 22, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=false2025.03 | 41.18 | — | |
| WEDGEEncoder=ResNet, No real data usage=false2023.12 | 41.1 | — | |
| WEDGE*Venue=ICRA 23, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=true2025.03 | 41.06 | — | |
| DIDEXEncoder=ResNet, No real data usage=true, UDA component=MIC [17]2023.12 | 40.9 | — | |
| DPCLBackbone=ResNet-50, External Dataset=false, External Module=true, Source Domain=GTAV2024.03 | 40.59 | — | |
| SPCSource Dataset=GTA5, Backbone=ResNet-502023.07 | 40.46 | — | |
| SPCBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=GTAV2024.03 | 40.46 | — | |
| GTREncoder=ResNet, No real data usage=false2023.12 | 39.6 | — | |
| GTR*Venue=TIP 21, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=true2025.03 | 39.6 | — | |
| SHADESource Dataset=GTA5, Backbone=ResNet-502023.07 | 39.28 | — | |
| AdvStyleVenue=NeurIPS 22, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=false2025.03 | 39.27 | — | |
| DIRLSource Dataset=GTA5, Backbone=ResNet-502023.07 | 39.15 | — | |
| DIRLBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=GTAV2024.03 | 39.15 | — | |
| RobustNetEncoder=ResNet, No real data usage=true2023.12 | 38.7 | — | |
| DRPCEncoder=ResNet, No real data usage=false2023.12 | 38.7 | — | |
| HRDAVenue=TPAMI 23, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=false2025.03 | 38.69 | — | |
| WildNetSource Dataset=GTA5, Backbone=ResNet-502023.07 | 38.42 | — | |
| WildNetBackbone=ResNet-50, External Dataset=true, External Module=false, Source Domain=GTAV2024.03 | 38.42 | — | |
| SiamDoGeBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=GTAV2024.03 | 37.54 | — | |
| SAWSource Dataset=GTA5, Backbone=ResNet-502023.07 | 37.34 | — | |
| SANSAWBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=GTAV2024.03 | 37.34 | — | |
| IBN-NetEncoder=ResNet, No real data usage=true2023.12 | 36.7 | — | |
| BaselineEncoder=ResNet, No real data usage=true2023.12 | 36.6 | — | |
| SWEncoder=ResNet, No real data usage=true2023.12 | 36.6 | — | |
| AdvStyleSource Dataset=GTA5, Backbone=ResNet-502023.07 | 35.54 | — | |
| ISWSource Dataset=GTA5, Backbone=ResNet-502023.07 | 35.2 | — | |
| RobustNetBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=GTAV2024.03 | 35.2 | — | |
| IBN-NetVenue=ECCV 18, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=false2025.03 | 34.21 | — | |
| GTRSource Dataset=GTA5, Backbone=ResNet-502023.07 | 33.75 | — | |
| ISWVenue=CVPR 21, Backbone=ResNet-101, Source Domain=GTA5, Use of additional real-world data=false2025.03 | 33.36 | — | |
| IternormSource Dataset=GTA5, Backbone=ResNet-502023.07 | 32.7 | — | |
| IBNSource Dataset=GTA5, Backbone=ResNet-502023.07 | 32.3 | — | |
| IBN-NetBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=GTAV2024.03 | 32.3 | — | |
| DRPCSource Dataset=GTA5, Backbone=ResNet-502023.07 | 32.14 | — | |
| IWSource Dataset=GTA5, Backbone=ResNet-502023.07 | 27.48 | — | |
| BaselineBackbone=ResNet-50, External Dataset=false, External Module=false, Source Domain=GTAV2024.03 | 25.14 | — |