We have already explored the architecture of the discriminator network in The architecture of the discriminator network section. Let's start by writing the layers for the discriminator network in the Keras framework and then create a Keras model, using the functional API of the Keras framework.

Perform the following steps to implement the discriminator network in Keras:

1. Start by defining the hyperparameters required for the discriminator network:

leakyrelu_alpha = 0.2
momentum = 0.8
input_shape = (256, 256, 3)

2. Next, create an input layer to feed input to the network, as follows:

input_layer = Input(shape=input_shape)

3. Next, add a convolution block, as follows:
   Configuration:
   • Filters: 64
   • Kernel size: 3
   • Strides: 1
   • Padding: same
   • Activation: LeakyReLU with alpha equal to 0.2:

dis1 = Conv2D(filters=64, kernel_size=3, strides=1, padding='same')(input_layer)
dis1 = LeakyReLU(alpha=leakyrelu_alpha)(dis1)

4. Next, add another seven convolution blocks, as follows:
   Configuration:
   • Filters: 64, 128, 128, 256, 256, 512, 512
   • Kernel size: 3, 3, 3, 3, 3, 3, 3
   • Strides: 2, 1, 2, 1, 2, 1, 2
   • Padding: same for each convolution layer
   • Activation: LealyReLU with alpha equal to 0.2 for each convolution layer:

# Add the 2nd convolution block
dis2 = Conv2D(filters=64, kernel_size=3, strides=2, padding='same')(dis1)
dis2 = LeakyReLU(alpha=leakyrelu_alpha)(dis2)
dis2 = BatchNormalization(momentum=momentum)(dis2)

# Add the third convolution block
dis3 = Conv2D(filters=128, kernel_size=3, strides=1, padding='same')(dis2)
dis3 = LeakyReLU(alpha=leakyrelu_alpha)(dis3)
dis3 = BatchNormalization(momentum=momentum)(dis3)

# Add the fourth convolution block
dis4 = Conv2D(filters=128, kernel_size=3, strides=2, padding='same')(dis3)
dis4 = LeakyReLU(alpha=leakyrelu_alpha)(dis4)
dis4 = BatchNormalization(momentum=0.8)(dis4)

# Add the fifth convolution block
dis5 = Conv2D(256, kernel_size=3, strides=1, padding='same')(dis4)
dis5 = LeakyReLU(alpha=leakyrelu_alpha)(dis5)
dis5 = BatchNormalization(momentum=momentum)(dis5)

# Add the sixth convolution block
dis6 = Conv2D(filters=256, kernel_size=3, strides=2, padding='same')(dis5)
dis6 = LeakyReLU(alpha=leakyrelu_alpha)(dis6)
dis6 = BatchNormalization(momentum=momentum)(dis6)

# Add the seventh convolution block
dis7 = Conv2D(filters=512, kernel_size=3, strides=1, padding='same')(dis6)
dis7 = LeakyReLU(alpha=leakyrelu_alpha)(dis7)
dis7 = BatchNormalization(momentum=momentum)(dis7)

# Add the eight convolution block
dis8 = Conv2D(filters=512, kernel_size=3, strides=2, padding='same')(dis7)
dis8 = LeakyReLU(alpha=leakyrelu_alpha)(dis8)
dis8 = BatchNormalization(momentum=momentum)(dis8)

5. Next, add a dense layer with 1,024 nodes, as follows:
   Configuration:
   • Nodes: 1024
   • Activation: LeakyReLU with alpha equal to 0.2:

dis9 = Dense(units=1024)(dis8)
dis9 = LeakyReLU(alpha=0.2)(dis9)

6. Then, add a dense layer to return the probabilities, as follows:

output = Dense(units=1, activation='sigmoid')(dis9)

7. Finally, create a Keras model and specify the inputs and the outputs for the network:

model = Model(inputs=[input_layer], outputs=[output], 
              name='discriminator')

Wrap the entire code for the discriminator network inside a function, as follows:

def build_discriminator():
    """
    Create a discriminator network using the hyperparameter values defined below
    :return:
    """
    leakyrelu_alpha = 0.2
    momentum = 0.8
    input_shape = (256, 256, 3)

    input_layer = Input(shape=input_shape)

    # Add the first convolution block
    dis1 = Conv2D(filters=64, kernel_size=3, strides=1, padding='same')(input_layer)
    dis1 = LeakyReLU(alpha=leakyrelu_alpha)(dis1)

    # Add the 2nd convolution block
    dis2 = Conv2D(filters=64, kernel_size=3, strides=2, padding='same')(dis1)
    dis2 = LeakyReLU(alpha=leakyrelu_alpha)(dis2)
    dis2 = BatchNormalization(momentum=momentum)(dis2)

    # Add the third convolution block
    dis3 = Conv2D(filters=128, kernel_size=3, strides=1, padding='same')(dis2)
    dis3 = LeakyReLU(alpha=leakyrelu_alpha)(dis3)
    dis3 = BatchNormalization(momentum=momentum)(dis3)

    # Add the fourth convolution block
    dis4 = Conv2D(filters=128, kernel_size=3, strides=2, padding='same')(dis3)
    dis4 = LeakyReLU(alpha=leakyrelu_alpha)(dis4)
    dis4 = BatchNormalization(momentum=0.8)(dis4)

    # Add the fifth convolution block
    dis5 = Conv2D(256, kernel_size=3, strides=1, padding='same')(dis4)
    dis5 = LeakyReLU(alpha=leakyrelu_alpha)(dis5)
    dis5 = BatchNormalization(momentum=momentum)(dis5)

    # Add the sixth convolution block
    dis6 = Conv2D(filters=256, kernel_size=3, strides=2, padding='same')(dis5)
    dis6 = LeakyReLU(alpha=leakyrelu_alpha)(dis6)
    dis6 = BatchNormalization(momentum=momentum)(dis6)

    # Add the seventh convolution block
    dis7 = Conv2D(filters=512, kernel_size=3, strides=1, padding='same')(dis6)
    dis7 = LeakyReLU(alpha=leakyrelu_alpha)(dis7)
    dis7 = BatchNormalization(momentum=momentum)(dis7)

    # Add the eight convolution block
    dis8 = Conv2D(filters=512, kernel_size=3, strides=2, padding='same')(dis7)
    dis8 = LeakyReLU(alpha=leakyrelu_alpha)(dis8)
    dis8 = BatchNormalization(momentum=momentum)(dis8)

    # Add a dense layer
    dis9 = Dense(units=1024)(dis8)
    dis9 = LeakyReLU(alpha=0.2)(dis9)

    # Last dense layer - for classification
    output = Dense(units=1, activation='sigmoid')(dis9)

    model = Model(inputs=[input_layer], outputs=[output], name='discriminator')
    return model

In this section, we have successfully created a Keras model for the discriminator network. In the next section, we will build the VGG19 network, as shown in the Introducing SRGAN section.