Code Implementation Of VGG-Net

from keras.layers import Input, Lambda, Dense, Flatten
from keras.models import Model
from keras.applications.vgg16 import VGG16
from keras.applications.vgg16 import preprocess_input
from keras.preprocessing import image
from keras.preprocessing.image import ImageDataGenerator
from keras.models import Sequential
import numpy as np
from glob import glob
import matplotlib.pyplot as plt

import warnings
warnings.filterwarnings("ignore", category=FutureWarning)

IMAGE_SIZE = [224, 224]

train_path = 'data/train'
test_path = 'data/test'

vgg = VGG16(input_shape=IMAGE_SIZE + [3], weights='imagenet', include_top=False)

for layer in vgg.layers:
  layer.trainable = False

folders = glob('data/train/*')
print(len(folders))

x = Flatten()(vgg.output)
prediction = Dense(len(folders), activation='softmax')(x)
model = Model(inputs=vgg.input, outputs=prediction)
model.summary()

from keras import optimizers


sgd = optimizers.SGD(lr=0.01, decay=1e-6, momentum=0.9)
model.compile(loss='binary_crossentropy',
              optimizer=sgd,
              metrics=['accuracy'])

train_datagen = ImageDataGenerator(
    preprocessing_function=preprocess_input,
    rotation_range=40,
    width_shift_range=0.2,
    height_shift_range=0.2,
    shear_range=0.2,
    zoom_range=0.2,
    horizontal_flip=True,
    fill_mode='nearest')

test_datagen = ImageDataGenerator(
    preprocessing_function=preprocess_input,
    rotation_range=40,
    width_shift_range=0.2,
    height_shift_range=0.2,
    shear_range=0.2,
    zoom_range=0.2,
    horizontal_flip=True,
    fill_mode='nearest')

from datetime import datetime
from keras.callbacks import ModelCheckpoint, LearningRateScheduler
from keras.callbacks import ReduceLROnPlateau

#lr_scheduler = LearningRateScheduler(lr_schedule)

lr_reducer = ReduceLROnPlateau(factor=np.sqrt(0.1),
                               cooldown=0,
                               patience=5,
                               min_lr=0.5e-6)

#num_epochs = 1000
#num_batch_size = 32

checkpoint = ModelCheckpoint(filepath='mymodel.h5', 
                               verbose=1, save_best_only=True)

callbacks = [checkpoint, lr_reducer, lr_scheduler]

start = datetime.now()

model.fit_generator(
  train_set,
  validation_data=test_set,
  epochs=10,
  steps_per_epoch=5,
  validation_steps=32,
    callbacks=callbacks ,verbose=1)


duration = datetime.now() - start
print("Training completed in time: ", duration)