Fig. 2
From: A review of convolutional neural networks in computer vision
Structure of CNN (Suppose this is an n-classification problem. The original data is convolved twice (Convolution 1, Convolution 2), pooled twice (Max Pooling 1, Max Pooling 2), and output to the fully connected layer (Fully connection), and finally the Softmax activation function compresses the output vectors of the full connection layer into (0, 1) and outputs them in the output layer. The Data Cost 1 represents the probability of belonging to the n categories; the larger the value, the greater the possibility of belonging to the category.)