Table 3 Neural networks and backbones complexity in terms of FLoating point Operations Per Second (FLOPS), number of parameters and layers.
From: A semi-automatic toolbox for markerless effective semantic feature extraction
Net/Backbone | FLOPS (bilions) | # Params (milions) | # Layers |
|---|---|---|---|
FPN/Efficientb1 | 12.80 | 0.96 | 379 |
FPN/ResNet50 | 6.43 | 2.69 | 237 |
LinkNet/Efficientb1 | 8.04 | 0.86 | 388 |
LinkNet/ResNet50 | 2.09 | 2.88 | 246 |
PSPNet/Efficientb1 | 1.76 | 0.18 | 142 |
PSPNet/ResNet50 | 0.90 | 0.39 | 116 |
Unet/Efficientb1 | 8.72 | 1.26 | 373 |
Unet/ResNet50 | 2.58 | 3.26 | 231 |
