Table 18 Inference metrics of FCAU-Net model.
From: Feature fusion context attention gate UNet for detection of polycystic ovary syndrome
Model | Inferenc time per image (ms) | Inferenc time per batch (ms) | Throughput (images/sec) | Latency (ms) | ||||
|---|---|---|---|---|---|---|---|---|
Raw images | FCE images | Raw images | FCE images | Raw images | FCE images | Raw images | FCE images | |
DenseNet | 12.5 | 12.1 | 120 | 115 | 8.0 | 8.3 | 12.5 | 12.1 |
VGG | 14.2 | 13.8 | 135 | 130 | 7.4 | 7.7 | 14.2 | 13.8 |
AlexNet | 10.4 | 10.1 | 100 | 95 | 9.6 | 9.9 | 10.4 | 10.1 |
ResNet | 11.8 | 11.4 | 110 | 108 | 9.1 | 9.3 | 11.8 | 11.4 |
U-Net | 15.6 | 15.0 | 150 | 145 | 6.7 | 7.0 | 15.6 | 15.0 |
Attention U-Net | 16.3 | 16.0 | 160 | 155 | 6.3 | 6.5 | 16.3 | 16.0 |
Proposed FCAU-Net | 9.1 | 8.7 | 90 | 85 | 11.1 | 11.5 | 9.1 | 8.7 |
