create_model.py

from tensorflow.keras import datasets, layers, models, optimizers, regularizers, callbacks
from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint

def create_model(image_shape=(256, 4096, 1), print_summary=False):
    # initial model
    model = models.Sequential()

    # CONV layer: filter 16, stride 7x7
    model.add(layers.Conv2D(16, (7, 7),input_shape=image_shape))
    # Batch Normalization layer -> avoid overfitting
    model.add(layers.BatchNormalization())
    # activation layer 
    model.add(layers.Activation('relu'))
    # max pooling -> reduce image size
    model.add(layers.MaxPooling2D((2, 2)))
    # droput later -> avoid overfitting
    model.add(layers.Dropout(0.25))

    model.add(layers.Conv2D(32, (5, 5), padding="same"))
    # Batch Normalization layer -> avoid overfitting
    model.add(layers.BatchNormalization())
    # activation layer 
    model.add(layers.Activation('relu'))
    # max pooling -> reduce image size
    model.add(layers.MaxPooling2D((2, 2)))
    # droput later -> avoid overfitting
    model.add(layers.Dropout(0.25))
    
    model.add(layers.Conv2D(64, (3, 3), padding="same"))
    # Batch Normalization layer -> avoid overfitting
    model.add(layers.BatchNormalization())
    # activation layer 
    model.add(layers.Activation('relu'))
    # max pooling -> reduce image size
    model.add(layers.MaxPooling2D((2, 2)))
    # droput later -> avoid overfitting
    model.add(layers.Dropout(0.25))
    
    model.add(layers.Conv2D(64, (3, 3), padding="same"))
    # Batch Normalization layer -> avoid overfitting
    model.add(layers.BatchNormalization())
    # activation layer 
    model.add(layers.Activation('relu'))
    # max pooling -> reduce image size
    model.add(layers.MaxPooling2D((2, 2)))
    # droput later -> avoid overfitting
    model.add(layers.Dropout(0.25))
    
    # flatten later -> from matrix to vector
    model.add(layers.Flatten())
    
    # fully connected layer -> nn layer with 64 nodes
    model.add(layers.Dense(64))
    # Batch Normalization layer -> avoid overfitting
    model.add(layers.BatchNormalization())
    # activation layer 
    model.add(layers.Activation('relu'))
    # droput later -> avoid overfitting
    model.add(layers.Dropout(0.25))

    # fully connected layer -> nn layer with 64 nodes
    model.add(layers.Dense(64))
    # Batch Normalization layer -> avoid overfitting
    model.add(layers.BatchNormalization())
    # activation layer 
    model.add(layers.Activation('relu'))
    # droput later -> avoid overfitting
    model.add(layers.Dropout(0.25))
    
    # output layer
    model.add(layers.Dense(1, activation='sigmoid'))
    
    # set model compiler
    model.compile(optimizer='SGD', loss='binary_crossentropy', metrics=['accuracy'])
    
    # show the CNN model detail
    if print_summary:
        model.summary()
    return model

def train_model(model, xtrain, ytrain, xval, yval, n_epoch, batch_size):
    # train CNN model
    # batch size to reduce memory usage
    # set early stopping to avoid overfitting
    
    earlystopping = EarlyStopping(monitor='val_accuracy', patience=2)
    filepath = project_path + "/model/weights-improvement-{epoch:02d}-{val_accuracy:.2f}.hdf5"
    
    checkpoint = ModelCheckpoint(filepath, monitor='val_accuracy', verbose=1, save_best_only=True, mode='max')
    callbacks_list = [checkpoint, earlystopping]

    history = model.fit(xtrain, ytrain, epochs=n_epoch, batch_size=batch_size, validation_data=(xval, yval), callbacks=[callbacks_list])
    return history

create_model(image_shape=(256, 4096, 1), print_summary=True)