keras_simple_model.py

from keras.layers import Input, Dense, Activation, ZeroPadding2D, BatchNormalization, Flatten, Conv2D
# Kept the unused imports for future reference
from keras.layers import AveragePooling2D, MaxPooling2D, Dropout, GlobalMaxPooling2D, GlobalAveragePooling2D
from keras.models import Model
import keras.backend as K
from IPython.display import SVG
from keras.utils.vis_utils import model_to_dot
from keras.utils import plot_model

K.set_image_data_format('channels_last')


def cnn_model(input_shape, show_table=False):
    """
    Implementation of the HappyModel.

    Arguments:
    input_shape -- shape of the images of the dataset

    Returns:
    model -- a Model() instance in Keras
    """

    X_input = Input(input_shape)
    X = ZeroPadding2D((3, 3))(X_input)

    X = Conv2D(32, (7, 7), strides=(1, 1), name='conv0')(X)
    X = BatchNormalization(axis=3, name='bn0')(X)
    X = Activation('relu')(X)
    X = MaxPooling2D((2, 2), name='max_pool')(X)

    X = Flatten()(X)
    X = Dense(1, activation='sigmoid', name='fc')(X)

    model = Model(inputs=X_input, outputs=X, name='HappyModel')

    if show_table:
        model.summary()

    return model


def show_model_graph(model_instance, file_name):
    plot_model(model_instance, to_file=file_name)
    SVG(model_to_dot(model_instance).create(prog='dot', format='svg'))