Instance Segmentation on ADE20K 15 (val)
37.8APOneFormer
Evaluation Results
| Method | Links | |
|---|---|---|
| OneFormerBackbone=Swin-L, #Params=219M, Crop Size=1280x1280, Extra Data=false, Training Protocol=Joint Training2022.11 | 37.8 | |
| OneFormerBackbone=Swin-L, #Params=219M, Crop Size=896x896, Extra Data=false, Training Protocol=Joint Training2022.11 | 37.6 | |
| OneFormerBackbone=DINAT-L, #Params=223M, Crop Size=1280x1280, Extra Data=false, Training Protocol=Joint Training2022.11 | 37.1 | |
| OneFormerBackbone=DINAT-L, #Params=223M, Crop Size=896x896, Extra Data=false, Training Protocol=Joint Training2022.11 | 36.8 | |
| OneFormerBackbone=ConvNeXt-XL, #Params=372M, Crop Size=640x640, Extra Data=false, Training Protocol=Joint Training2022.11 | 36.3 | |
| OneFormerBackbone=ConvNeXt-L, #Params=220M, Crop Size=640x640, Extra Data=false, Training Protocol=Joint Training2022.11 | 36.2 | |
| OneFormerBackbone=DINAT-L, #Params=223M, Crop Size=640x640, Extra Data=false, Training Protocol=Joint Training2022.11 | 36 | |
| OneFormerBackbone=Swin-L, #Params=219M, Crop Size=640x640, Extra Data=false, Training Protocol=Joint Training2022.11 | 35.9 | |
| Mask2Former-InstanceBackbone=Swin-L, #Params=216M, Crop Size=640x640, Extra Data=false, Training Protocol=Individual Training2022.11 | 34.9 | |
| Mask2Former-PanopticBackbone=Swin-L, #Params=216M, Crop Size=640x640, Extra Data=false, Training Protocol=Individual Training2022.11 | 34.2 |