Table 19 FCAU-Net model complexity and efficiency metrics.
From: Feature fusion context attention gate UNet for detection of polycystic ovary syndrome
Model | Number of parameters (M) | FLOPS ( X 109) | Computation overhead | |||
|---|---|---|---|---|---|---|
Raw images | FCE images | Raw images | FCE images | Raw images | FCE images | |
DenseNet | 8.1 | 8.1 | 5.6 | 5.5 | Moderate | Moderate |
VGG | 14.7 | 14.7 | 7.8 | 7.7 | High | High |
AlexNet | 5.6 | 5.6 | 3.5 | 3.4 | Low | Low |
ResNet | 11.2 | 11.2 | 6.7 | 6.6 | Moderate | Moderate |
U-Net | 12.8 | 12.8 | 9.4 | 9.3 | High | High |
Attention U-Net | 14.2 | 14.2 | 10.2 | 10.1 | High | High |
Proposed FCAU-Net | 7.3 | 7.3 | 4.9 | 4.8 | Low | Low |
