TJU暑期的深度学习训练营，这是人脸识别运用图像增强后的一段代码~

import os, shutil
# ! unzip tjudataset.zip

base_dir = './tjudataset'

# read data
train_dir = os.path.join(base_dir,'train')
validation_dir = os.path.join(base_dir,'validation')
test_dir = os.path.join(base_dir,'test')

from keras import layers
from keras import models

from keras import optimizers

from keras.preprocessing.image import ImageDataGenerator

model = models.Sequential()
model.add(layers.Conv2D(64, (2, 2), activation='relu',
                        input_shape=(210, 210, 3)))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(128, (2, 2), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(256, (2, 2), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(512, (2, 2), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Flatten())
model.add(layers.Dropout(0.3))
model.add(layers.Dense(512, activation='relu'))
model.add(layers.Dropout(0.2))
model.add(layers.Dense(61, activation='softmax'))

model.compile(loss='categorical_crossentropy',
              optimizer=optimizers.RMSprop(lr=1e-4),
              metrics=['acc'])
model.summary()

_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
conv2d_1 (Conv2D)            (None, 209, 209, 64)      832       
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 104, 104, 64)      0         
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 103, 103, 128)     32896     
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 51, 51, 128)       0         
_________________________________________________________________
conv2d_3 (Conv2D)            (None, 50, 50, 256)       131328    
_________________________________________________________________
max_pooling2d_3 (MaxPooling2 (None, 25, 25, 256)       0         
_________________________________________________________________
conv2d_4 (Conv2D)            (None, 24, 24, 512)       524800    
_________________________________________________________________
max_pooling2d_4 (MaxPooling2 (None, 12, 12, 512)       0         
_________________________________________________________________
flatten_1 (Flatten)          (None, 73728)             0         
_________________________________________________________________
dropout_1 (Dropout)          (None, 73728)             0         
_________________________________________________________________
dense_1 (Dense)              (None, 512)               37749248  
_________________________________________________________________
dropout_2 (Dropout)          (None, 512)               0         
_________________________________________________________________
dense_2 (Dense)              (None, 61)                31293     
=================================================================
Total params: 38,470,397
Trainable params: 38,470,397
Non-trainable params: 0
_________________________________________________________________

train_datagen = ImageDataGenerator(
    rescale=1./255,
    rotation_range=10,
    width_shift_range=0.2,
    height_shift_range=0.2,
    shear_range=0.2,
    zoom_range=0.2,
    horizontal_flip=True,
    fill_mode='nearest')

# Note that the validation data should not be augmented!
validation_datagen = ImageDataGenerator(rescale=1./255)

train_generator = train_datagen.flow_from_directory(
        # This is the target directory
        train_dir,
        # All images will be resized to 150x150
        target_size=(210, 210),
        batch_size=61,
        # Since we use binary_crossentropy loss, we need binary labels
        class_mode='categorical')

validation_generator = validation_datagen.flow_from_directory(
        validation_dir,
        target_size=(210, 210),
        batch_size=61,
        class_mode='categorical')

Found 549 images belonging to 61 classes.
Found 61 images belonging to 61 classes.

from keras.callbacks import ModelCheckpoint   
from matplotlib import pyplot as plt
import numpy as np
checkpointer = ModelCheckpoint(filepath='TJUFACE.augmentation.model.weights.best.hdf5', verbose=1, 
                               save_best_only=True)
before = 0
test_datagen = ImageDataGenerator(rescale=1./255)
test_generator = test_datagen.flow_from_directory(
        test_dir,
        target_size=(210, 210),
        batch_size=61,
        class_mode='categorical')
xx = []
yy = []

for i in range(200):
    xx += [i]
    print('The ',i+1,'times:')
    history = model.fit_generator(
      train_generator,
      steps_per_epoch=9,
      epochs=1, 
      validation_data=validation_generator,
      validation_steps=1, 
      callbacks=[checkpointer], 
      verbose=1)
#     model.load_weights('TJUFACE.augmentation.model.weights.best.hdf5')
    test_loss, test_acc = model.evaluate_generator(test_generator, steps=1)
    yy += [test_acc]
    if test_acc > before:
        print('-------------------------------------------------------------------------')
        print('epochs = ',i+1)
        print('Test_acc:', test_acc)
        print('-------------------------------------------------------------------------')
        before = test_acc
print()
print('The highest test_acc :',before)

Found 61 images belonging to 61 classes.
The  1 times:
Epoch 1/1
9/9 [==============================] - 14s 2s/step - loss: 4.1443 - acc: 0.0109 - val_loss: 4.0953 - val_acc: 0.0984

Epoch 00001: val_loss improved from inf to 4.09529, saving model to TJUFACE.augmentation.model.weights.best.hdf5
-------------------------------------------------------------------------
epochs =  1
Test_acc: 0.09836065769195557
-------------------------------------------------------------------------
The  2 times:
Epoch 1/1
9/9 [==============================] - 11s 1s/step - loss: 4.1036 - acc: 0.0328 - val_loss: 4.0612 - val_acc: 0.0656

Epoch 00001: val_loss improved from 4.09529 to 4.06118, saving model to TJUFACE.augmentation.model.weights.best.hdf5
The  3 times:
Epoch 1/1
9/9 [==============================] - 11s 1s/step - loss: 4.0701 - acc: 0.0510 - val_loss: 3.9763 - val_acc: 0.0656
……
……
……
Epoch 00001: val_loss did not improve from 0.33823
The  199 times:
Epoch 1/1
9/9 [==============================] - 11s 1s/step - loss: 0.4557 - acc: 0.8743 - val_loss: 0.8803 - val_acc: 0.8361

Epoch 00001: val_loss did not improve from 0.33823
The  200 times:
Epoch 1/1
9/9 [==============================] - 11s 1s/step - loss: 0.5764 - acc: 0.8452 - val_loss: 0.4710 - val_acc: 0.8852

Epoch 00001: val_loss did not improve from 0.33823

The highest test_acc : 0.9508196711540222

plt.figure(figsize=(10,5),dpi=200)
plt.title('Test_acc')
plt.xlabel('Epochs')
plt.xticks(np.arange(0,205,10))
plt.ylabel('Test_acc')
plt.yticks(np.arange(0,1,0.1))
def smooth_curve(points, factor=0.8):
  smoothed_points = []
  for point in points:
    if smoothed_points:
      previous = smoothed_points[-1]
      smoothed_points.append(previous * factor + point * (1 - factor))
    else:
      smoothed_points.append(point)
  return smoothed_points
plt.plot(xx,smooth_curve(yy),label='Final',color='m',marker=',',linestyle='-')# 16进制颜色码
plt.legend()
plt.show()

model.load_weights('TJUFACE.augmentation.model.weights.best.hdf5')

test_datagen = ImageDataGenerator(rescale=1./255)
test_generator = test_datagen.flow_from_directory(
        test_dir,
        target_size=(210, 210),
        batch_size=61,
        class_mode='categorical')

test_loss, test_acc = model.evaluate_generator(test_generator, steps=1)
print('test acc:', test_acc)

Found 61 images belonging to 61 classes.
test acc: 0.9016393423080444