MobileNet Model

import numpy as np
import os
import numpy
from tensorflow.python.keras.layers import Dense
from tensorflow.python.keras import optimizers
from keras import regularizers
from keras.regularizers import l2
from keras.layers import Dropout
from keras.applications.mobilenet import MobileNet
from keras.layers import GlobalAveragePooling2D, Dense, Dropout, Flatten, BatchNormalization
from keras.models import Sequential
from keras.preprocessing.image import ImageDataGenerator
from keras.callbacks import EarlyStopping, ModelCheckpoint, LearningRateScheduler
from keras.optimizers import Adam
import sys
from keras import optimizers

# Fixed for our classes
NUM_CLASSES = 4
# Fixed for color images
CHANNELS = 3

IMAGE_RESIZE = 224
DENSE_LAYER_ACTIVATION = 'softmax'
OBJECTIVE_FUNCTION = 'categorical_crossentropy'

LOSS_METRICS = ['accuracy']
NUM_EPOCHS = 1000
EARLY_STOP_PATIENCE = 50


BATCH_SIZE_TRAINING = 128
BATCH_SIZE_VALIDATION = 64

STEPS_PER_EPOCH_TRAINING = 5096/BATCH_SIZE_TRAINING
STEPS_PER_EPOCH_VALIDATION = 1456/BATCH_SIZE_VALIDATION

base_mobilenet_model = MobileNet(include_top = False, weights = None)
model = Sequential()
model.add(Dense(3,input_shape = [IMAGE_RESIZE,IMAGE_RESIZE,3]))
model.add(base_mobilenet_model)
model.add(Dropout(0.5))
model.add(BatchNormalization())
model.add(GlobalAveragePooling2D())

# 2nd layer as Dense for 4-class classification,
model.add(Dense(NUM_CLASSES, activation = DENSE_LAYER_ACTIVATION))

model.summary()

opt = optimizers.SGD(lr=0.001, decay=1e-6, momentum=0.5, nesterov=True)
model.compile(optimizer = opt, loss = OBJECTIVE_FUNCTION,  metrics =  LOSS_METRICS)

image_size = IMAGE_RESIZE
shift = 0.2
data_generator = ImageDataGenerator(rescale=1.0/255.0,
                                   width_shift_range=shift,
                                   height_shift_range=shift,
                                   horizontal_flip=True,
                                   vertical_flip=True,
                                   rotation_range=90,
                                   brightness_range=[0.2,1.0],
                                   zoom_range=[0.5,1.0])

data_generator2 = ImageDataGenerator(rescale=1.0/255.0)

train_generator = data_generator.flow_from_directory(
        'trainset/',
        target_size=(image_size, image_size),
        batch_size=BATCH_SIZE_TRAINING,
        class_mode='categorical')

validation_generator = data_generator2.flow_from_directory(
        'testset/',
        target_size=(image_size, image_size),
        batch_size=BATCH_SIZE_VALIDATION,
        class_mode='categorical')

filepath="weights-improvement-{epoch:02d}-vacc:{val_accuracy:.2f}-tacc:{accuracy:.2f}.hdf5"

cb_early_stopper = EarlyStopping(monitor = 'val_accuracy', mode='max', verbose=1, patience = EARLY_STOP_PATIENCE)
cb_checkpointer = ModelCheckpoint(filepath = filepath, monitor = 'val_accuracy', save_best_only = False, mode = 'auto')

fit_history = model.fit_generator(train_generator,
        steps_per_epoch=STEPS_PER_EPOCH_TRAINING,
        epochs = NUM_EPOCHS,
        validation_data=validation_generator,
        validation_steps=STEPS_PER_EPOCH_VALIDATION,
        verbose=2,
        callbacks = [cb_checkpointer, cb_early_stopper]
)