Continual Semantic Segmentation on ADE20k 100-10 (6 tasks) (val)
0.3722mIoU (101-150)offline
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| offlineModel=Mask2Former2026.01 | 0.3722 | — | 0.4487 | — | 0.4869 | |
| Joint2023.05 | 0.282 | 0.443 | 0.389 | — | — | |
| CPP+Model=Mask2Former2026.01 | 0.2222 | — | 0.3625 | — | 0.4327 | |
| FairCLBackbone=SegFormer2023.05 | 0.2191 | 0.4221 | 0.3549 | 39.36 | — | |
| MiB + AWT2023.05 | 0.2128 | 0.391 | 0.332 | — | — | |
| FairCLBackbone=DeepLab-V32023.05 | 0.2036 | 0.4173 | 0.3465 | 39.01 | — | |
| CPPModel=MaskFormer2026.01 | 0.2035 | — | 0.3478 | — | 0.4199 | |
| SATS2023.05 | 0.1909 | 0.4142 | 0.3418 | — | — | |
| SSUL2023.05 | 0.1875 | 0.402 | 0.331 | — | — | |
| LAGModel=DeepLabv32026.01 | 0.1869 | — | 0.3356 | — | 0.41 | |
| IDECModel=DeepLabv32026.01 | 0.1762 | — | 0.3271 | — | 0.4025 | |
| RCIL2023.05 | 0.176 | 0.393 | 0.321 | — | — | |
| RCILModel=DeepLabv32026.01 | 0.176 | — | 0.321 | — | 0.393 | |
| PLOPBackbone=ResNet-101, Architecture=Deeplab-V3, Pre-training=ImageNet, CSS Setting=Overlapped2020.11 | 0.1361 | 40.48 | 0.3159 | 0.3664 | — | |
| PLOP2021.06 | 0.1361 | 0.4048 | 0.3159 | 0.3664 | — | |
| PLOP2023.05 | 0.1361 | 0.4048 | 0.3159 | 36.64 | — | |
| PLOPModel=DeepLabv32026.01 | 0.1361 | — | 0.3159 | — | 0.4048 | |
| MiBBackbone=ResNet-101, Architecture=Deeplab-V3, Pre-training=ImageNet, CSS Setting=Overlapped2020.11 | 0.1112 | 38.21 | 0.2924 | 0.3512 | — | |
| MiB2021.06 | 0.1112 | 0.3821 | 0.2924 | 0.3512 | — | |
| MiB2023.05 | 0.1112 | 0.3821 | 0.2924 | 35.12 | — | |
| MiBModel=DeepLabv32026.01 | 0.1112 | — | 0.2924 | — | 0.3821 | |
| ILTBackbone=ResNet-101, Architecture=Deeplab-V3, Pre-training=ImageNet, CSS Setting=Overlapped2020.11 | 0.0306 | 0.11 | 0.0109 | 0.1256 | — | |
| ILT2021.06 | 0.0306 | 0.0011 | 0.0109 | 0.1256 | — | |
| ILT2023.05 | 0.0306 | 0.0011 | 0.0109 | 12.56 | — |