CNN之从头训练一个猫狗图片分类模型

猫狗图片下载地址：

链接：https://pan.baidu.com/s/1l1AnBgkAAEhh0vI5_loWKw
提取码：2xq4
说明：大概有816M大小，分为train和test，train有cat和dog标签作为图片名字。

一、建立训练、验证、测试图片集

import os
import shutil

original_dataset_dir = "/home/suanfa/picture/dogs-vs-cats/train/train"
base_dir = "dogs_and_cats_small"
os.makedirs(base_dir)

## TODO 建立训练、验证、测试文件夹
train_dir = os.path.join(base_dir, 'train')
os.makedirs(train_dir)
validation_dir = os.path.join(base_dir, 'validation')
os.makedirs(validation_dir)
test_dir = os.path.join(base_dir, 'test')
os.makedirs(test_dir)

## TODO 建立猫狗训练文件夹
train_cats_dir = os.path.join(train_dir, 'cats')
os.makedirs(train_cats_dir)
train_dogs_dir = os.path.join(train_dir, 'dogs')
os.makedirs(train_dogs_dir)

## TODO 建立猫狗验证文件夹
validation_cats_dir = os.path.join(validation_dir, 'cats')
os.makedirs(validation_cats_dir)
validation_dogs_dir = os.path.join(validation_dir, 'dogs')
os.makedirs(validation_dogs_dir)

## TODO 建立猫狗测试文件夹
test_cats_dir = os.path.join(test_dir, 'cats')
os.makedirs(test_cats_dir)
test_dogs_dir = os.path.join(test_dir, 'dogs')
os.makedirs(test_dogs_dir)

## TODO 1000张猫训练图片
fnames = ['cat.{}.jpg'.format(i) for i in range(1000)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(train_cats_dir, fname)
    shutil.copyfile(src, dst)

## TODO 500张猫验证图片
fnames = ['cat.{}.jpg'.format(i) for i in range(1000, 1500)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(validation_cats_dir, fname)
    shutil.copyfile(src, dst)

## TODO 500张猫测试证图片
fnames = ['cat.{}.jpg'.format(i) for i in range(1500, 2000)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(test_cats_dir, fname)
    shutil.copyfile(src, dst)

## TODO 1000张狗训练图片
fnames = ['dog.{}.jpg'.format(i) for i in range(1000)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(train_dogs_dir, fname)
    shutil.copyfile(src, dst)

## TODO 1000张狗验证图片
fnames = ['dog.{}.jpg'.format(i) for i in range(1000, 1500)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(validation_dogs_dir, fname)
    shutil.copyfile(src, dst)

## TODO 1000张狗测试图片
fnames = ['dog.{}.jpg'.format(i) for i in range(1500, 2000)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(test_dogs_dir, fname)
    shutil.copyfile(src, dst)

print("total training cat images: ", len(os.listdir(train_cats_dir)))
print("total training dog images: ", len(os.listdir(train_dogs_dir)))
print("total validation cat images: ", len(os.listdir(validation_cats_dir)))
print("total validation dog images: ", len(os.listdir(validation_dogs_dir)))
print("total test cat images: ", len(os.listdir(test_cats_dir)))
print("total test dog images: ", len(os.listdir(test_dogs_dir)))

运行的结果为：
total training cat images: 1000
total training dog images: 1000
total validation cat images: 500
total validation dog images: 500
total test cat images: 500
total test dog images: 500

二、构建CNN模型

from keras import layers
from keras import models
from keras import optimizers

model = models.Sequential()
model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(150, 150, 3)))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))

model.add(layers.Flatten())
model.add(layers.Dense(512, activation='relu'))
model.add(layers.Dense(1, activation='sigmoid'))

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

三、读取和处理图像

from keras.preprocessing.image import ImageDataGenerator

train_datagen = ImageDataGenerator(rescale=1./255)  # 像素归一
test_datagen = ImageDataGenerator(rescale=1./255)

train_generator = train_datagen.flow_from_directory(
    "./dogs_and_cats_small/train",
    target_size=(150,150),
    batch_size=20,
    class_mode='binary'
)
validation_generator = test_datagen.flow_from_directory(
    "./dogs_and_cats_small/validation",
    target_size=(150, 150),
    batch_size=20,
    class_mode='binary'
)
for data_batch, labels_batch in train_generator:
     print("data batch shape: ", data_batch.shape)
     print("labels batch shape: ", labels_batch.shape)
    break

打印出来结果为：

Found 2000 images belonging to 2 classes.
Found 1000 images belonging to 2 classes.
data batch shape: (20, 150, 150, 3)
labels batch shape: (20,)

四、训练和保存模型

history = model.fit_generator(
    train_generator,
    steps_per_epoch=100,
    epochs=30,
    validation_data=validation_generator,
    validation_steps=50
)
model.save('cats_and_dogs_small_1.h5')

五、绘制训练和验证的精度和损失曲线

import matplotlib.pyplot as plt
acc = history.history['acc']
val_acc = history.history['val_acc']
loss = history.history['loss']
val_loss = history.history['val_loss']

epochs = range(1, len(acc)+1)

plt.plot(epochs, acc, 'bo', label='Training acc')
plt.plot(epochs, val_acc, 'b', label='Validation acc')
plt.title('Training and Validation accuracy')
plt.legend()

plt.figure()

plt.plot(epochs, loss, 'bo', label='Training loss')
plt.plot(epochs, val_loss, 'b', label='Validation loss')
plt.title('Training and Validation loss')
plt.legend()

plt.show()

训练和验证精度曲线图：

训练和验证损失曲线图：

六、结论

从以上图片可看出：
训练精度随时间线性增加，直到接近100%，而验证精度则停留在70%-72%
验证损失在10个epochs后达到最小，之后开始上升，而训练损失一直线性下降，直至接近0
因为仅有2000个训练样本，出现了过拟合，可以使用降低过拟合的方法：L2正则化或dropout，
然而，在用深度学习模型处理图像几乎总是用到：数据增强。

七、参考资料

• https://www.cnblogs.com/xiximayou/p/12372969.html#top
• 佛朗索瓦.肖莱著，张亮译. Python深度学习[M]. 人民邮电出版社.2018.8.p102-111.

END