Research Article
Multiactivation Pooling Method in Convolutional Neural Networks for Image Recognition
Table 10
The number of trainable parameters and depth in different VGG models are shown in Table 1. “Conv-D” indicates the number of convolutional layers and “FC-D” indicates the number of fully connected layers.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||