Table 8 Resource metrics results of feature extraction models with DNN classification for skin disease detection and classification.
From: Skin disease diagnostics through federated transfer learning on heterogeneous data
Minimum | Maximum | Average | Minimum | Maximum | Average | |
|---|---|---|---|---|---|---|
Model | GPU memory used (%) | GPU process used (%) | ||||
DenseNet | 66.250 | 71.450 | 68.850 | 58.360 | 63.580 | 60.970 |
VGG19 | 65.850 | 70.680 | 68.265 | 58.120 | 62.750 | 60.435 |
Xception | 66.480 | 71.120 | 68.800 | 59.450 | 63.250 | 61.350 |
UNet | 67.120 | 71.880 | 69.500 | 60.120 | 64.120 | 62.120 |
CPU process used (%) | Virtual memory used (%) | |||||
DenseNet | 7.550 | 9.250 | 8.400 | 75.360 | 78.450 | 76.905 |
VGG19 | 7.250 | 8.950 | 8.100 | 74.650 | 77.850 | 76.250 |
Xception | 7.650 | 9.350 | 8.500 | 75.580 | 78.950 | 77.265 |
UNet | 7.850 | 9.580 | 8.715 | 76.120 | 79.150 | 77.635 |
