第十六节，使用函数封装库tf.contrib.layers
这一节，介绍TensorFlow中的一个封装好的高级库，里面有前面讲过的很多函数的高级封装，使用这个高级库来开发程序将会提高效率。
我们改写第十三节的程序，卷积函数我们使用tf.contrib.layers.conv2d()，池化函数使用tf.contrib.layers.max_pool2d()和tf.contrib.layers.avg_pool2d()，全连接函数使用tf.contrib.layers.fully_connected()。
一 tf.contrib.layers中的具体函数介绍
1.tf.contrib.layers.conv2d()函数的定义如下：
def convolution(inputs,
num_outputs,
kernel_size,
stride=1,
padding='SAME',
data_format=None,
rate=1,
activation_fn=nn.relu,
normalizer_fn=None,
normalizer_params=None,
weights_initializer=initializers.xavier_initializer(),
weights_regularizer=None,
biases_initializer=init_ops.zeros_initializer(),
biases_regularizer=None,
reuse=None,
variables_collections=None,
outputs_collections=None,
trainable=True,
scope=None):
常用的参数说明如下：
- inputs:形状为[batch_size, height, width, channels]的输入。
- num_outputs：代表输出几个channel。这里不需要再指定输入的channel了，因为函数会自动根据inpus的shpe去判断。
- kernel_size：卷积核大小，不需要带上batch和channel，只需要输入尺寸即可。[5,5]就代表5x5的卷积核，如果长和宽都一样，也可以只写一个数5.
- stride：步长，默认是长宽都相等的步长。卷积时，一般都用1，所以默认值也是1.如果长和宽都不相等，也可以用一个数组[1,2]。
- padding：填充方式，'SAME'或者'VALID'。
- activation_fn：激活函数。默认是ReLU。也可以设置为None
- weights_initializer：权重的初始化，默认为initializers.xavier_initializer()函数。
- weights_regularizer：权重正则化项，可以加入正则函数。biases_initializer：偏置的初始化，默认为init_ops.zeros_initializer()函数。
- biases_regularizer：偏置正则化项，可以加入正则函数。
- trainable：是否可训练，如作为训练节点，必须设置为True，默认即可。如果我们是微调网络，有时候需要冻结某一层的参数，则设置为False。
2.tf.contrib.layers.max_pool2d()函数的定义如下：
def max_pool2d(inputs,
kernel_size,
stride=2,
padding='VALID',
data_format=DATA_FORMAT_NHWC,
outputs_collections=None,
scope=None):
参数说明如下：
- inputs: A 4-D tensor of shape `[batch_size, height, width, channels]` if`data_format` is `NHWC`, and `[batch_size, channels, height, width]` if `data_format` is `NCHW`.
- kernel_size: A list of length 2: [kernel_height, kernel_width] of the pooling kernel over which the op is computed. Can be an int if both values are the same.
- stride: A list of length 2: [stride_height, stride_width].Can be an int if both strides are the same. Note that presently both strides must have the same value.
- padding: The padding method, either 'VALID' or 'SAME'.
- data_format: A string. `NHWC` (default) and `NCHW` are supported.
- outputs_collections: The collections to which the outputs are added.
- scope: Optional scope for name_scope.
3.tf.contrib.layers.avg_pool2d()函数定义
def avg_pool2d(inputs,
kernel_size,
stride=2,
padding='VALID',
data_format=DATA_FORMAT_NHWC,
outputs_collections=None,
scope=None):
参数说明如下：
- inputs: A 4-D tensor of shape `[batch_size, height, width, channels]` if`data_format` is `NHWC`, and `[batch_size, channels, height, width]` if `data_format` is `NCHW`.
- kernel_size: A list of length 2: [kernel_height, kernel_width] of the pooling kernel over which the op is computed. Can be an int if both values are the same.
- stride: A list of length 2: [stride_height, stride_width].Can be an int if both strides are the same. Note that presently both strides must have the same value.
- padding: The padding method, either 'VALID' or 'SAME'.
- data_format: A string. `NHWC` (default) and `NCHW` are supported.
- outputs_collections: The collections to which the outputs are added.
- scope: Optional scope for name_scope.
4.tf.contrib.layers.fully_connected()函数的定义如下：
def fully_connected(inputs,
num_outputs,
activation_fn=nn.relu,
normalizer_fn=None,
normalizer_params=None,
weights_initializer=initializers.xavier_initializer(),
weights_regularizer=None,
biases_initializer=init_ops.zeros_initializer(),
biases_regularizer=None,
reuse=None,
variables_collections=None,
outputs_collections=None,
trainable=True,
scope=None):
参数说明如下：
- inputs: A tensor of at least rank 2 and static value for the last dimension; i.e. `[batch_size, depth]`, `[None, None, None, channels]`.
- num_outputs: Integer or long, the number of output units in the layer.
- activation_fn: Activation function. The default value is a ReLU function.Explicitly set it to None to skip it and maintain a linear activation.
- normalizer_fn: Normalization function to use instead of `biases`. If `normalizer_fn` is provided then `biases_initializer` and
- `biases_regularizer` are ignored and `biases` are not created nor added.default set to None for no normalizer function
- normalizer_params: Normalization function parameters.
- weights_initializer: An initializer for the weights.
- weights_regularizer: Optional regularizer for the weights.
- biases_initializer: An initializer for the biases. If None skip biases.
- biases_regularizer: Optional regularizer for the biases.
- reuse: Whether or not the layer and its variables should be reused. To be able to reuse the layer scope must be given.
- variables_collections: Optional list of collections for all the variables or a dictionary containing a different list of collections per variable.
- outputs_collections: Collection to add the outputs.
- trainable: If `True` also add variables to the graph collection `GraphKeys.TRAINABLE_VARIABLES` (see tf.Variable).如果我们是微调网络，有时候需要冻结某一层的参数，则设置为False。
- scope: Optional scope for variable_scope.
二 改写cifar10分类
代码如下：
# -*- coding: utf-8 -*-
"""
Created on Thu May 3 12:29:16 2018
@author: zy
"""
'''
建立一个带有全连接层的卷积神经网络 并对CIFAR-10数据集进行分类
1.使用2个卷积层的同卷积操作，滤波器大小为5x5，每个卷积层后面都会跟一个步长为2x2的池化层，滤波器大小为2x2
2.对输出的64个feature map进行全局平均池化，得到64个特征
3.加入一个全连接层，使用softmax激活函数，得到分类
'''
import cifar10_input
import tensorflow as tf
import numpy as np
def print_op_shape(t):
'''
输出一个操作op节点的形状
'''
print(t.op.name,'',t.get_shape().as_list())
'''
一 引入数据集
'''
batch_size = 128
learning_rate = 1e-4
training_step = 15000
display_step = 200
#数据集目录
data_dir = './cifar10_data/cifar-10-batches-bin'
print('begin')
#获取训练集数据
images_train,labels_train = cifar10_input.inputs(eval_data=False,data_dir = data_dir,batch_size=batch_size)
print('begin data')
'''
二 定义网络结构
'''
#定义占位符
input_x = tf.placeholder(dtype=tf.float32,shape=[None,24,24,3]) #图像大小24x24x
input_y = tf.placeholder(dtype=tf.float32,shape=[None,10]) #0-9类别
x_image = tf.reshape(input_x,[batch_size,24,24,3])
#1.卷积层 ->池化层
h_conv1 = tf.contrib.layers.conv2d(inputs=x_image,num_outputs=64,kernel_size=5,stride=1,padding='SAME', activation_fn=tf.nn.relu) #输出为[-1,24,24,64]
print_op_shape(h_conv1)
h_pool1 = tf.contrib.layers.max_pool2d(inputs=h_conv1,kernel_size=2,stride=2,padding='SAME') #输出为[-1,12,12,64]
print_op_shape(h_pool1)
#2.卷积层 ->池化层
h_conv2 =tf.contrib.layers.conv2d(inputs=h_pool1,num_outputs=64,kernel_size=[5,5],stride=[1,1],padding='SAME', activation_fn=tf.nn.relu) #输出为[-1,12,12,64]
print_op_shape(h_conv2)
h_pool2 = tf.contrib.layers.max_pool2d(inputs=h_conv2,kernel_size=[2,2],stride=[2,2],padding='SAME') #输出为[-1,6,6,64]
print_op_shape(h_pool2)
#3全连接层
nt_hpool2 = tf.contrib.layers.avg_pool2d(inputs=h_pool2,kernel_size=6,stride=6,padding='SAME') #输出为[-1,1,1,64]
print_op_shape(nt_hpool2)
nt_hpool2_flat = tf.reshape(nt_hpool2,[-1,64])
y_conv = tf.contrib.layers.fully_connected(inputs=nt_hpool2_flat,num_outputs=10,activation_fn=tf.nn.softmax)
print_op_shape(y_conv)
'''
三 定义求解器
'''
#softmax交叉熵代价函数
cost = tf.reduce_mean(-tf.reduce_sum(input_y * tf.log(y_conv),axis=1))
#求解器
train = tf.train.AdamOptimizer(learning_rate).minimize(cost)
#返回一个准确度的数据
correct_prediction = tf.equal(tf.arg_max(y_conv,1),tf.arg_max(input_y,1))
#准确率
accuracy = tf.reduce_mean(tf.cast(correct_prediction,dtype=tf.float32))
'''
四 开始训练
'''
sess = tf.Session();
sess.run(tf.global_variables_initializer())
# 启动计算图中所有的队列线程 调用tf.train.start_queue_runners来将文件名填充到队列，否则read操作会被阻塞到文件名队列中有值为止。
tf.train.start_queue_runners(sess=sess)
for step in range(training_step):
#获取batch_size大小数据集
image_batch,label_batch = sess.run([images_train,labels_train])
#one hot编码
label_b = np.eye(10,dtype=np.float32)[label_batch]
#开始训练
train.run(feed_dict={input_x:image_batch,input_y:label_b},session=sess)
if step % display_step == 0:
train_accuracy = accuracy.eval(feed_dict={input_x:image_batch,input_y:label_b},session=sess)
print('Step {0} tranining accuracy {1}'.format(step,train_accuracy))
