Deep convolutional generative adversarial networks

The deep convolutional generative adversarial networks (DCGAN) are introduced in the paper: Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks, by A. Radford, L. Metz, and S. Chintala, arXiv: 1511.06434, 2015. The generator uses a 100-dimensional, uniform distribution space, Z, which is then projected into a smaller space by a series of vis-a-vis convolution operations. An example is shown in the following figure:

A DCGAN generator can be described by the following Keras code; it is also described by one implementation, available at: https://github.com/jacobgil/keras-dcgan:

def generator_model():
    model = Sequential()
    model.add(Dense(input_dim=100, output_dim=1024))
    model.add(Activation('tanh'))
    model.add(Dense(128*7*7))
    model.add(BatchNormalization())
    model.add(Activation...