Table 4 Layer wise parameters.
Methods | Trainable parameters weights(M) | Precision | Recall | F1-Score | IoU |
|---|---|---|---|---|---|
U-Net n = 3(n-depth of U-Net) | 9.5 | 0.831 | 0.834 | 0.837 | 0.839 |
U-Net n = 4(n-depth of U-Net) | 10 | 0.845 | 0.847 | 0.849 | 0.841 |
U-Net n = 5(n-depth of U-Net) | 11 | 0.851 | 0.852 | 0.854 | 0.859 |
U-Net + Spatial Attention | 11.5 | 0.891 | 0.894 | 0.889 | 0.895 |
Deep Shallow U-Net + Spatial Attention | 4 | 0.95 | 0.951 | 0.953 | 0.954 |
