input layer S-layer V-layer C-layer S-layer V-layer C-layer S-layer V-layer C-layer S-layer V-layer C-layer
Begin Previous page Next page Contents Index

5. Network structure - Layers


Each stage of the neocognitron consists of certain number of layers of given type.

Structure of the network from figure 4.1 after drawing of layers of which the particular stages are assembled is shown in figure 5.1.

Fig. 5.1 - Network structure - Layers
Fig. 5.1 - Network structure - Layers

From figure 5.1 it is obvious that four types of layers exist in the neocognitron. Stage 0 always consists of only one input layer. All higher stages consist of one S-layer, one V-layer and one C-layer.

In figure 5.1 we have also established ordinarily used notation of layers in the neocognitron. We will use this notation, described in table 5.1, in the following text as well.

SymbolDenotes
U0input layer
USlS-layer in the l-th stage of the network
UVlV-layer in the l-th stage of the network
UClC-layer in the l-th stage of the network
Tab. 5.1 - Notation used for layers in the neocognitron

Begin Previous page Next page Contents Index