Semantic Segmentation on PASCAL VOC 2012 x8 (test)
63.3mIoUIHR -> T
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| IHR -> TSR Method=Oracle (IHR), Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 63.3 | 0 | — | |
| SR4IRSR Method=SwinIR, Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 56.5 | 0.355 | 24.23 | |
| SR4IRSR Method=EDSR-baseline, Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 55 | 0.38 | 23.91 | |
| S -> TSR Method=SwinIR, Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 53.9 | 0.387 | 24.75 | |
| S -> TSR Method=EDSR-baseline, Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 52.3 | 0.407 | 24.41 | |
| S + TSR Method=EDSR-baseline, Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 50 | 0.48 | 23.58 | |
| S + TSR Method=SwinIR, Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 50 | 0.48 | 23.63 | |
| ILR -> TSR Method=Bilinear up, Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 49.3 | 0.476 | 22.54 | |
| T -> SSR Method=EDSR-baseline, Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 49.2 | 0.461 | 23.14 | |
| T -> SSR Method=SwinIR, Task Network=DeepLabV3, Backbone=MobileNetV32024.04 | 48.7 | 0.466 | 23.06 |