利用TensorFlow和Keras搭配卷积神经网络之猫狗数据集分类实验
目录
一、环境配置
1、安装anaconda
对于anaconda的安装,大家可以参考博客:
Anaconda安装教程&粗解
2、Jupyter notebook配置
修改运行工作目录
打开菜单目录,修改快捷键的属性即可
修改目标为你的工作空间
界面简单介绍
添加jupyter_conrib_nbextension插件
原生jupyter是没有:自动补全代码功能+pep8+字体大小+代码行号+拼写检查+目录索引功能的,需要插件帮助我们完成
①安装jupyter_contrib_nbextensions库
pip install jupyter_contrib_nbextensions -i https://pypi.douban.com/simple/
②配置到jupyter
jupyter contrib nbextension install --user --skip-running-check
③重启jupyter notebook
安装完成后,重新启动jupyter notebook,“Nbextensions”出现在导航栏中,勾选目录(勾选方框内容)
3、配置TensorFlow、Keras
在cmd中pip安装
pip install tensorflow
pip install keras
二、猫狗数据集分类
1、猫狗数据集下载
百度网盘链接:
猫狗数据集
提取码:33ff
下载解压后出现两个文件夹
将解压后的猫狗数据集放在新建的文件夹中,并重新命名为dogs-vs-cats
2、数据集分类
分类代码如下
import tensorflow as tf
import keras
import os, shutil
# 原始目录所在的路径
original_dataset_dir = 'E:\mg\dogs-vs-cats\train'
# 数据集分类后的目录
base_dir = 'E:\mg\dogs-vs-cats-sort'
os.mkdir(base_dir)
# # 训练、验证、测试数据集的目录
train_dir = os.path.join(base_dir, 'train')
os.mkdir(train_dir)
validation_dir = os.path.join(base_dir, 'validation')
os.mkdir(validation_dir)
test_dir = os.path.join(base_dir, 'test')
os.mkdir(test_dir)
# 猫训练图片所在目录
train_cats_dir = os.path.join(train_dir, 'cats')
os.mkdir(train_cats_dir)
# 狗训练图片所在目录
train_dogs_dir = os.path.join(train_dir, 'dogs')
os.mkdir(train_dogs_dir)
# 猫验证图片所在目录
validation_cats_dir = os.path.join(validation_dir, 'cats')
os.mkdir(validation_cats_dir)
# 狗验证数据集所在目录
validation_dogs_dir = os.path.join(validation_dir, 'dogs')
os.mkdir(validation_dogs_dir)
# 猫测试数据集所在目录
test_cats_dir = os.path.join(test_dir, 'cats')
os.mkdir(test_cats_dir)
# 狗测试数据集所在目录
test_dogs_dir = os.path.join(test_dir, 'dogs')
os.mkdir(test_dogs_dir)
# 将前1000张猫图像复制到train_cats_dir
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)
# 将下500张猫图像复制到validation_cats_dir
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)
# 将下500张猫图像复制到test_cats_dir
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)
# 将前1000张狗图像复制到train_dogs_dir
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)
# 将下500张狗图像复制到validation_dogs_dir
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)
# 将下500张狗图像复制到test_dogs_dir
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)
显示对于目录下相片数量
#输出数据集对应目录下图片数量
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)))
三、搭配卷积神经网络
搭配网络模型
#网络模型构建
from keras import layers
from keras import models
#keras的序贯模型
model = models.Sequential()
#卷积层,卷积核是3*3,激活函数relu
model.add(layers.Conv2D(32, (3, 3), activation='relu',
input_shape=(150, 150, 3)))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#卷积层,卷积核2*2,激活函数relu
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#卷积层,卷积核是3*3,激活函数relu
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#卷积层,卷积核是3*3,激活函数relu
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#flatten层,用于将多维的输入一维化,用于卷积层和全连接层的过渡
model.add(layers.Flatten())
#全连接,激活函数relu
model.add(layers.Dense(512, activation='relu'))
#全连接,激活函数sigmoid
model.add(layers.Dense(1, activation='sigmoid'))
了解征图尺寸变化
model.summary()
使用二元交叉矩阵作为我们的损失
from keras import optimizers
model.compile(loss='binary_crossentropy',
optimizer=optimizers.RMSprop(lr=1e-4),
metrics=['acc'])
数据预处理
from keras.preprocessing.image import ImageDataGenerator
# All images will be rescaled by 1./255
train_datagen = ImageDataGenerator(rescale=1./255)
test_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=(150, 150),
batch_size=20,
# Since we use binary_crossentropy loss, we need binary labels
class_mode='binary')
validation_generator = test_datagen.flow_from_directory(
validation_dir,
target_size=(150, 150),
batch_size=20,
class_mode='binary')
查看生成器输出
for data_batch, labels_batch in train_generator:
print('data batch shape:', data_batch.shape)
print('labels batch shape:', labels_batch.shape)
break
使生成器模型适用于数据
history = model.fit_generator(
train_generator,
steps_per_epoch=100,
epochs=30,
validation_data=validation_generator,
validation_steps=50)
结果可视化
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(len(acc))
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()
四、数据基准调整
图像数据生成器增强数据
datagen = ImageDataGenerator(
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')
- rotation_range是一个角度值(0-180),在这个范围内可以随机旋转图片。
- width_shift和height_shift是范围(作为总宽度或高度的一部分),在其中可以随机地垂直或水平地转换图片。
- shear_range用于随机应用剪切转换。
- zoom_range用于在图片内部随机缩放。
- horizontal_flip是用于水平随机翻转一半的图像——当没有假设水平不对称时(例如真实世界的图片)。
- fill_mode是用于填充新创建像素的策略,它可以在旋转或宽度/高度移动之后出现。
查看增强后的图像
# This is module with image preprocessing utilities
from keras.preprocessing import image
fnames = [os.path.join(train_cats_dir, fname) for fname in os.listdir(train_cats_dir)]
# We pick one image to "augment"
img_path = fnames[3]
# Read the image and resize it
img = image.load_img(img_path, target_size=(150, 150))
# Convert it to a Numpy array with shape (150, 150, 3)
x = image.img_to_array(img)
# Reshape it to (1, 150, 150, 3)
x = x.reshape((1,) + x.shape)
# The .flow() command below generates batches of randomly transformed images.
# It will loop indefinitely, so we need to `break` the loop at some point!
i = 0
for batch in datagen.flow(x, batch_size=1):
plt.figure(i)
imgplot = plt.imshow(image.array_to_img(batch[0]))
i += 1
if i % 4 == 0:
break
plt.show()
模型中增加一个Dropout层
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.Dropout(0.5))
model.add(layers.Dense(512, activation='relu'))
model.add(layers.Dense(1, activation='sigmoid'))
model.compile(loss='binary_crossentropy',
optimizer=optimizers.RMSprop(lr=1e-4),
metrics=['acc'])
用数据增强和退出来训练我们的网络
train_datagen = ImageDataGenerator(
rescale=1./255,
rotation_range=40,
width_shift_range=0.2,
height_shift_range=0.2,
shear_range=0.2,
zoom_range=0.2,
horizontal_flip=True,)
# Note that the validation data should not be augmented!
test_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=(150, 150),
batch_size=32,
# Since we use binary_crossentropy loss, we need binary labels
class_mode='binary')
validation_generator = test_datagen.flow_from_directory(
validation_dir,
target_size=(150, 150),
batch_size=32,
class_mode='binary')
history = model.fit_generator(
train_generator,
steps_per_epoch=100,
epochs=100,
validation_data=validation_generator,
validation_steps=50)
数据增强效果
数据增强和dropout层增加效果
可视化
acc = history.history['acc']
val_acc = history.history['val_acc']
loss = history.history['loss']
val_loss = history.history['val_loss']
epochs = range(len(acc))
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()
数据增强和dropout层增加的可视化结果
五、Vgg19完成猫狗分类
初始化一个VGG19网络实例
from keras.applications import VGG19
conv_base = VGG19(weights = 'imagenet',include_top = False,input_shape=(150, 150, 3))
conv_base.summary()
首次运行时候,会自动从对应网站下载h5格式文件
建议使用浏览器下载
from keras.applications import VGG19
conv_base = VGG16(weights = 'vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5', include_top = False, input_shape=(150, 150, 3))
conv_base.summary()
结果如下
将猫狗数据集传递给神经网络
import os
import numpy as np
from keras.preprocessing.image import ImageDataGenerator
base_dir = 'E:\mg\dogs-vs-cats-sort'
train_dir = os.path.join(base_dir, 'train')
validation_dir = os.path.join(base_dir, 'validation')
test_dir = os.path.join(base_dir, 'test')
datagen = ImageDataGenerator(rescale = 1. / 255)
batch_size = 20
def extract_features(directory, sample_count):
features = np.zeros(shape = (sample_count, 4, 4, 512))
labels = np.zeros(shape = (sample_count))
generator = datagen.flow_from_directory(directory, target_size = (150, 150),
batch_size = batch_size,
class_mode = 'binary')
i = 0
for inputs_batch, labels_batch in generator:
#把图片输入VGG16卷积层,让它把图片信息抽取出来
features_batch = conv_base.predict(inputs_batch)
#feature_batch 是 4*4*512结构
features[i * batch_size : (i + 1)*batch_size] = features_batch
labels[i * batch_size : (i+1)*batch_size] = labels_batch
i += 1
if i * batch_size >= sample_count :
#for in 在generator上的循环是无止境的,因此我们必须主动break掉
break
return features , labels
#extract_features 返回数据格式为(samples, 4, 4, 512)
train_features, train_labels = extract_features(train_dir, 2000)
validation_features, validation_labels = extract_features(validation_dir, 1000)
test_features, test_labels = extract_features(test_dir, 1000)
抽取的特征输入到我们自己的神经层中进行分类
from keras import models
from keras import layers
from keras import optimizers
#构造我们自己的网络层对输出数据进行分类
model = models.Sequential()
model.add(layers.Dense(256, activation='relu', input_dim = 4 * 4 * 512))
model.add(layers.Dropout(0.5))
model.add(layers.Dense(1, activation = 'sigmoid'))
model.compile(optimizer=optimizers.RMSprop(lr = 2e-5), loss = 'binary_crossentropy', metrics = ['acc'])
history = model.fit(train_features, train_labels, epochs = 30, batch_size = 20,
validation_data = (validation_features, validation_labels))
结果可视化
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 = 'Train_acc')
plt.plot(epochs, val_acc, 'b', label = 'Validation acc')
plt.title('Trainning 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()
六、相关解释及参考资料
1、相关解释
- 什么是overfit(过拟合)?
① 简单理解就是训练样本的得到的输出和期望输出基本一致,但是测试样本输出和测试样本的期望输出相差却很大 。
②为了得到一致假设而使假设变得过度复杂称为过拟合。
如果数据本身呈现二次型,故用一条二次曲线拟合会更好。但普通的PLS程序只提供线性方程供拟合之用。这就产生拟合不足即“欠拟合”现象,从而在预报时要造成偏差。如果我们用人工神经网络拟合,则因为三层人工神经网络拟合能力极强,有能力拟合任何函数。如果拟合彻底,就会连实验数据点分布不均匀,实验数据的误差等等“噪声”都按最小二乘判据拟合进数学模型。这当然也会造成预报的偏差。
- 什么是数据增强?
①数据集增强主要是为了减少网络的过拟合现象,通过对训练图片进行变换可以得到泛化能力更强的网络,更好的适应应用场景。
②数据增强也叫数据扩增,意思是在不实质性的增加数据的情况下,让有限的数据产生等价于更多数据的价值。
③常用的数据增强方法有:
- 旋转 | 反射变换(Rotation/reflection): 随机旋转图像一定角度; 改变图像内容的朝向;
- 翻转变换(flip): 沿着水平或者垂直方向翻转图像;
- 缩放变换(zoom): 按照一定的比例放大或者缩小图像;
- 平移变换(shift): 在图像平面上对图像以一定方式进行平移;
可以采用随机或人为定义的方式指定平移范围和平移步长, 沿水平或竖直方向进行平移. 改变图像内容的位置; - 尺度变换(scale): 对图像按照指定的尺度因子, 进行放大或缩小; 或者参照SIFT特征提取思想, 利用指定的尺度因子对图像滤波构造尺度空间. 改变图像内容的大小或模糊程度;
- 对比度变换(contrast): 在图像的HSV颜色空间,改变饱和度S和V亮度分量,保持色调H不变. 对每个像素的S和V分量进行指数运算(指数因子在0.25到4之间), 增加光照变化;
- 噪声扰动(noise): 对图像的每个像素RGB进行随机扰动, 常用的噪声模式是椒盐噪声和高斯噪声;
- 颜色变化: 在图像通道上添加随机扰动。
2、参考资料
基于jupyter notebook的python编程-----猫狗数据集的阶段分类得到模型精度并进行数据集优化
【TensorFlow&Keras】入门猫狗数据集实验–理解卷积神经网络CNN