Research Article
Proposing a Recognition System of Gestures Using MobilenetV2 Combining Single Shot Detector Network for Smart-Home Applications
Table 1
Comparing metrics of convolution networks for ImageNet dataset.
| No | Network | Accuracy on ImageNet | Number of parameter | Size of network (megabyte) | Depth |
| 1 | VGG-16 [5] | 0.901 | 138,357,544 | 528 | 23 | 2 | VGG-19 [16] | 0.90 | 143,667,240 | 549 | 26 | 3 | Inception-V3 [17] | 0.937 | 23,581,784 | 92 | 159 | 4 | Resnet-101 [15] | 0.938 | 44,675,560 | 171 | 101 | 5 | MobilenetV2 [6] | 0.901 | 3,538,984 | 14 | 88 | 6 | Densenet201 [18] | 0.936 | 20,242,984 | 80 | 201 | 7 | Xception [19] | 0.945 | 22,910,480 | 88 | 126 |
|
|