Neural Network Programming with TensorFlow

First convolution layer

After reshaping the input tensor into x_image, we will create the first convolution layer:

layer_conv1, weights_conv1 = new_conv_layer(input=x_image,
                                            num_input_channels=num_channels,
                                            filter_size=filter_size1,
                                            num_filters=num_filters1,
                                            use_pooling=True)
print(layer_conv1)

The new_conv_layer(...) function is defined here. Let's look at the value of each variable being sent to this function:

def new_conv_layer(input, # The previous layer. num_input_channels, # Num. channels in prev. layer. filter_size, # Width and height of each filter. num_filters, # Number of filters. use_pooling=True): # Use 2x2 max-pooling. # Shape ...