1.老师的源码
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt
import os
from tensorflow.examples.tutorials.mnist import input_data
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
# 控制训练过程的参数
learning_rate = 0.01
training_epochs = 20
batch_size = 128
display_step = 1
examples_to_show = 10
# 网络模型参数
n_hidden_units = 128 # 隐藏层神经元数量(让编码器和解码器都有同样规模的隐藏层)
n_input_units = 784 # 输入层神经元数量MNIST data input (img shape: 28*28)
n_output_units = n_input_units #解码器输出层神经元数量必须等于输入数据的units数量
#对一个张量进行全面的汇总(均值,标准差,最大最小值,直方图)
# (用于 TensorBoard 可视化).(www.studyai.com)
def variable_summaries(var):
with tf.name_scope('summaries'):
mean = tf.reduce_mean(var)
tf.summary.scalar('mean', mean)
with tf.name_scope('stddev'):
stddev = tf.sqrt(tf.reduce_mean(tf.square(var - mean)))
tf.summary.scalar('stddev', stddev)
tf.summary.scalar('max', tf.reduce_max(var))
tf.summary.scalar('min', tf.reduce_min(var))
tf.summary.histogram('histogram', var)
#根据输入输出节点数量返回初始化好的指定名称的权重Variable
def WeightsVariable(n_in, n_out, name_str):
return tf.Variable(tf.random_normal([n_in, n_out]), dtype=tf.float32, name=name_str)
#根据输出节点数量返回初始化好的指定名称的偏置 Variable
def BiasesVariable(n_out, name_str):
return tf.Variable(tf.random_normal([n_out]), dtype=tf.float32, name=name_str)
# 构建编码器
def Encoder(x_origin, activate_func=tf.nn.sigmoid):
# 编码器第一隐藏层(www.studyai.com)
with tf.name_scope('Layer'):
weights = WeightsVariable(n_input_units, n_hidden_units, 'weights')
biases = BiasesVariable(n_hidden_units, 'biases')
x_code = activate_func(tf.add(tf.matmul(x_origin, weights), biases))
variable_summaries(weights)
variable_summaries(biases)
return x_code
# 构建解码器
def Decoder(x_code, activate_func=tf.nn.sigmoid):
# 解码器第一隐藏层(www.studyai.com)
with tf.name_scope('Layer'):
weights = WeightsVariable(n_hidden_units, n_output_units, 'weights')
biases = BiasesVariable(n_output_units, 'biases')
x_decode = activate_func(tf.add(tf.matmul(x_code, weights), biases))
variable_summaries(weights)
variable_summaries(biases)
return x_decode
#调用上面写的函数构造计算图
with tf.Graph().as_default():
# 计算图输入
with tf.name_scope('X_Origin'):
X_Origin = tf.placeholder(tf.float32, [None, n_input_units])
# 构建编码器模型
with tf.name_scope('Encoder'):
X_code = Encoder(X_Origin, activate_func=tf.nn.sigmoid)
# 构建解码器模型(www.studyai.com)
with tf.name_scope('Decoder'):
X_decode = Decoder(X_code, activate_func=tf.nn.sigmoid)
# 定义损失节点:重构数据与原始数据的误差平方和损失
with tf.name_scope('Loss'):
Loss = tf.reduce_mean(tf.pow(X_Origin - X_decode, 2))
# 定义优化器,训练节点
with tf.name_scope('Train'):
Optimizer = tf.train.RMSPropOptimizer(learning_rate)
Train = Optimizer.minimize(Loss)
# 为计算图添加损失节点的标量汇总(scalar summary).(www.studyai.com)
with tf.name_scope('LossSummary'):
tf.summary.scalar('loss', Loss)
tf.summary.scalar('learning_rate', learning_rate)
#为计算图添加图像汇总(image summary).
with tf.name_scope('image_summaries'):
image_original = tf.reshape(X_Origin, [-1, 28, 28, 1])
image_reconstructed = tf.reshape(X_decode, [-1, 28, 28, 1])
tf.summary.image('image_original', image_original, 9)
tf.summary.image('image_reconstructed', image_reconstructed, 9)
# 聚合所有汇总节点(www.studyai.com)
merged_summaries = tf.summary.merge_all()
# 为所有变量添加初始化节点
Init = tf.global_variables_initializer()
print('把计算图写入事件文件,在TensorBoard里面查看')
summary_writer = tf.summary.FileWriter(logdir='logs', graph=tf.get_default_graph())
summary_writer.flush()
# 导入 MNIST data
mnist = input_data.read_data_sets("../MNIST_data/", one_hot=True)
# # 产生会话session,启动计算图(www.studyai.com)
with tf.Session() as sess:
sess.run(Init)
total_batch = int(mnist.train.num_examples / batch_size)
# 训练指定轮数,每一轮包含若干个批次
for epoch in range(training_epochs):
# 每一轮(回合)都要把所有的batch跑一边(www.studyai.com)
for i in range(total_batch):
batch_xs, batch_ys = mnist.train.next_batch(batch_size)
# 运行优化器Train节点(backprop) 和 Loss节点 (获取损失值)
_, loss = sess.run([Train, Loss], feed_dict={X_Origin: batch_xs})
# 每一轮训完之后,输出logs
if epoch % display_step == 0:
print("Epoch:", '%04d' % (epoch + 1), "loss=", "{:.9f}".format(loss))
# 调用sess.run()方法运行汇总节点,更新事件文件.
summary_str = sess.run(merged_summaries, feed_dict={X_Origin: batch_xs})
summary_writer.add_summary(summary_str, epoch)
summary_writer.flush()
# 关闭summary_writer(www.studyai.com)
summary_writer.close()
print("模型训练完毕!")
# 把训练好的编码器-解码器模型用在测试集上,输出重建后的样本数据
reconstructions = sess.run(X_decode, feed_dict={X_Origin: mnist.test.images[:examples_to_show]})
# 比较原始图像与重建后的图像
f, a = plt.subplots(2, 10, figsize=(10, 2))
for i in range(examples_to_show):
a[0][i].imshow(np.reshape(mnist.test.images[i], (28, 28)))
a[1][i].imshow(np.reshape(reconstructions[i], (28, 28)))
f.show()
plt.draw()
plt.waitforbuttonpress()
2.关于压缩率和激活函数
通过老师的两个小实验发现,压缩后的结果数改成128后,解码后的效果变差,但是能区分数字。把激活函数修改成softplus后,效果比较差。因此就此案例来说,sotfplus不适用。
通过学习课程,明白了这个是不是图像压缩和解压的原理?