Figure 2. Overall architecture of the Visual Geometry Group—16 (VGG-16) model.

VGG-16 comprises five blocks and three fully connected layers. Each block comprises some convolutional layers followed by a max-pooling layer. After flattening the output matrix after block 5, there are two fully connected layers for binary classification. The DNN used ImageNet parameters as the default weights of blocks 1–4 (Nagasato et al., 2018).