自编码器是一种数据的压缩算法，其中数据的压缩和解压缩函数是数据相关的、从样本中训练而来的。大部分自编码器中，压缩和解压缩的函数是通过神经网络实现的。

1. 一个简单的自编码器

• 使用的是tensorflow自带的mnist数据集(灰度图，像素范围0~1)

    import numpy as np
    import tensorflow as tf
    import matplotlib.pyplot as plt
    from tensorflow.examples.tutorials.mnist import input_data
    mnist = input_data.read_data_sets('MNIST_data', validation_size=0)
  

• 搭建网络

    # Size of the encoding layer (the hidden layer)
    encoding_dim = 32
    image_size = mnist.train.images.shape[1] #28x28=784
    inputs_ = tf.placeholder(tf.float32, (None, image_size), name='inputs')
    targets_ = tf.placeholder(tf.float32, (None, image_size), name='targets')
    # Output of hidden layer
    encoded = tf.layers.dense(inputs_, encoding_dim, activation=tf.nn.relu)
    # Output layer logits
    logits = tf.layers.dense(encoded, image_size, activation=None)
    # Sigmoid output from
    decoded = tf.nn.sigmoid(logits, name='output')
    # loss and optimization
    loss = tf.nn.sigmoid_cross_entropy_with_logits(labels=targets_, logits=logits)
    cost = tf.reduce_mean(loss)
    opt = tf.train.AdamOptimizer(0.001).minimize(cost)
  

• 训练网络

    # Create the session
    sess = tf.Session()
    epochs = 20
    batch_size = 200
    sess.run(tf.global_variables_initializer())
    for e in range(epochs):
        for ii in range(mnist.train.num_examples//batch_size):
            batch = mnist.train.next_batch(batch_size)
            feed = {inputs_: batch[0], targets_: batch[0]}
            batch_cost, _ = sess.run([cost, opt], feed_dict=feed)
            print("Epoch: {}/{}...".format(e+1, epochs), "Training loss: {:.4f}".format(batch_cost))
  

• 检验网络

    fig, axes = plt.subplots(nrows=2, ncols=10, sharex=True, sharey=True, figsize=(20,4))
    in_imgs = mnist.test.images[:10]
    reconstructed, compressed = sess.run([decoded, encoded], feed_dict={inputs_: in_imgs})
    # plot
    for images, row in zip([in_imgs, reconstructed], axes):
        for img, ax in zip(images, row):
            ax.imshow(img.reshape((28, 28)), cmap='Greys_r')
            ax.get_xaxis().set_visible(False)
            ax.get_yaxis().set_visible(False)
    fig.tight_layout(pad=0.1)
    sess.close()
  

2. 使用CNN搭建自编码器

• 搭建网络

    inputs_ = tf.placeholder(tf.float32, (None, 28, 28, 1), name='inputs')
    targets_ = tf.placeholder(tf.float32, (None, 28, 28, 1), name='targets')
    ### Encoder
    conv1 = tf.layers.conv2d(inputs_, 16, (3,3), padding='same', activation=tf.nn.relu) # 28x28x16
    maxpool1 = tf.layers.max_pooling2d(conv1, (2,2), (2,2), padding='same') # 14x14x16
    conv2 = tf.layers.conv2d(maxpool1, 8, (3,3), padding='same', activation=tf.nn.relu) # 14x14x8
    maxpool2 = tf.layers.max_pooling2d(conv2, (2,2), (2,2), padding='same') # 7x7x8
    conv3 = tf.layers.conv2d(maxpool2, 8, (3,3), padding='same', activation=tf.nn.relu) # 7x7x8
    encoded = tf.layers.max_pooling2d(conv3, (2,2), (2,2), padding='same') # 4x4x8
    ### Decoder
    upsample1 = tf.image.resize_nearest_neighbor(encoded, (7,7)) # 7x7x8
    conv4 = tf.layers.conv2d(upsample1, 8, (3,3), padding='same', activation=tf.nn.relu) # 7x7x8
    upsample2 = tf.image.resize_nearest_neighbor(conv4, (14,14)) # 14x14x8
    conv5 = tf.layers.conv2d(upsample2, 8, (3,3), padding='same', activation=tf.nn.relu) # 14x14x8
    upsample3 = tf.image.resize_nearest_neighbor(conv5, (28,28)) # 28x28x8
    conv6 = tf.layers.conv2d(upsample3, 16, (3,3), padding='same', activation=tf.nn.relu) # 28x28x16
    logits = tf.layers.conv2d(conv6, 1, (3,3), padding='same', activation=None) # 28x28x1
    decoded = tf.nn.sigmoid(logits, name='decoded') # 28x28x1
    ### Loss and Optimization:
    loss = tf.nn.sigmoid_cross_entropy_with_logits(labels=targets_, logits=logits)
    cost = tf.reduce_mean(loss)
    opt = tf.train.AdamOptimizer(0.001).minimize(cost)
  

  • 注意模型在解码部分使用的是upsample+convolution而不是transposed convolution(Reference)
• 训练网络

    sess = tf.Session()
    epochs = 20
    batch_size = 200
    sess.run(tf.global_variables_initializer())
    for e in range(epochs):
        for ii in range(mnist.train.num_examples//batch_size):
            batch = mnist.train.next_batch(batch_size)
            imgs = batch[0].reshape((-1, 28, 28, 1))
            batch_cost, _ = sess.run([cost, opt], feed_dict={inputs_: imgs, targets_: imgs})
            print("Epoch: {}/{}...".format(e+1, epochs), "Training loss: {:.4f}".format(batch_cost))
  

• 检验网络

    ###同上但注意数组维数
    reconstructed, compressed = sess.run([decoded, encoded], feed_dict={inputs_: in_imgs.reshape((10, 28, 28, 1))})
  

3. 使用自编码器降噪

• 搭建网络
  • 同上但feature map的个数由16-8-8-8-8-16变为32-32-16-16-32-32
• 训练网络

    sess = tf.Session()
    epochs = 100
    batch_size = 200
    # Set's how much noise we're adding to the MNIST images
    noise_factor = 0.5
    sess.run(tf.global_variables_initializer())
    for e in range(epochs):
        for ii in range(mnist.train.num_examples//batch_size):
            batch = mnist.train.next_batch(batch_size)
            # Get images from the batch
            imgs = batch[0].reshape((-1, 28, 28, 1))
            # Add random noise to the input images
            noisy_imgs = imgs + noise_factor * np.random.randn(*imgs.shape)
            # Clip the images to be between 0 and 1
            noisy_imgs = np.clip(noisy_imgs, 0., 1.)       
            # Noisy images as inputs, original images as targets
            batch_cost, _ = sess.run([cost, opt], feed_dict={inputs_: noisy_imgs, targets_: imgs})
            print("Epoch: {}/{}...".format(e+1, epochs), "Training loss: {:.4f}".format(batch_cost))
  

• 检验网络

    fig, axes = plt.subplots(nrows=2, ncols=10, sharex=True, sharey=True, figsize=(20,4))
    in_imgs = mnist.test.images[:10]
    noisy_imgs = in_imgs + noise_factor * np.random.randn(*in_imgs.shape)
    noisy_imgs = np.clip(noisy_imgs, 0., 1.)
    reconstructed = sess.run(decoded, feed_dict={inputs_: noisy_imgs.reshape((10, 28, 28, 1))})
    for images, row in zip([noisy_imgs, reconstructed], axes):
        for img, ax in zip(images, row):
            ax.imshow(img.reshape((28, 28)), cmap='Greys_r')
            ax.get_xaxis().set_visible(False)
            ax.get_yaxis().set_visible(False)
    fig.tight_layout(pad=0.1)
    sess.close()