“…(Bold represents optimal results with 100% training accuracy). Network | Image size | Layers | Connections | Train accuracy (%) | Test accuracy (%) | Training time (s) |
AlexNet [ 51 ] | 227 | 25 | 24 | 8.84 | 8.90 | 285 |
GoogLeNet [ 52 ] | 299 | 170 | 181 | 27.07 | 32.62 | 1563 |
Place365GoogLeNet | 224 | 144 | 170 | 13.81 | 14.50 | 1396 |
SqueezeNet [ 53 ] | 227 | 69 | 76 | 8.84 | 8.90 | 963 |
VGG16 [ 54 ] | 224 | 41 | 40 | 75.74 | 79.94 | 4237 |
VGG19 [ 55 ] | 224 | 47 | 46 | 53.16 | 50.25 | 4958 |
DarkNet53 [ 56 ] | 256 | 184 | 206 | 99.44 | 96.38 | 4319 |
InceptionV3 [ 57 ] | 299 | 315 | 349 | |
…”