Semantic Segmentation on ADE20K Circular Shift (val)
0.4442mIoUA-SwinV2-T
Evaluation results
| Method | Links | ||
|---|---|---|---|
| A-SwinV2-TBackbone=A-SwinV2-T, Segmenter=UperNet, Training Iterations=160K, Batch Size=162023.05 | 0.4442 | 1 | |
| SwinV2-TBackbone=SwinV2-T, Segmenter=UperNet, Training Iterations=160K, Batch Size=162023.05 | 0.4386 | 0.8816 | |
| A-Swin-TBackbone=A-Swin-T, Segmenter=UperNet, Training Iterations=160K, Batch Size=162023.05 | 0.4344 | 1 | |
| Swin-TBackbone=Swin-T, Segmenter=UperNet, Training Iterations=160K, Batch Size=162023.05 | 0.4293 | 0.8732 |