Udacity Deep Learning课程作业（四）

第四章作业主要是关于卷积神经网络的练习，涉及到tf.nn.conv2d、tf.nn.max_pooling等的使用。

相关API的用法

tf.nn.conv2d

tf.nn.conv2d(input, filter, strides, padding, use_cudnn_on_gpu=None, name=None)：

• input和filter都是4D的tensor类型，计算两者的2D卷积，input的shape为[batch, in_height, in_width, in_channels]，filter/kernel的shape为[filter_height, filter_width, in_channels, out_channels]
• strides为长度为4的列表，input的每个维度的滑动窗口的stride步长
• padding，使用的padding算法，”SAME”或者”VALID”，VALID卷积后比原来小，SAME卷积后和原来一样
• 步骤：
  
  • Flattens the filter to a 2-D matrix with shape [filter_height * filter_width * in_channels, output_channels].
  • Extracts image patches from the the input tensor to form a virtual tensor of shape [batch, out_height, out_width, filter_height * filter_width * in_channels].
  • For each patch, right-multiplies the filter matrix and the image patch vector.

output[b, i, j, k] =
    sum_{di, dj, q} input[b, strides[1] * i + di, strides[2] * j + dj, q] *
                    filter[di, dj, q, k]

测试代码：

import tensorflow as tf

image = tf.Variable(tf.random_normal([10, 12, 12, 3]))
conv = tf.Variable(tf.random_normal([3, 3, 3, 7]))
with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    print(sess.run(tf.nn.conv2d(image, conv, strides=[1, 5, 5, 1], padding="VALID")))

tf.nn.max_pool

max_pool(value, ksize, strides, padding, data_format='NHWC', name=None)：

• value: 4D tensor，shape为 [batch, height, width, channels]
• ksize: 长度大于等于4的列表，表示输入tensor在各个维度的窗口size
• strides: 长度大于等于4的列表，表示输入tensor在各个维度的滑动步长stride
• padding: ‘VALID’或者’SAME’
• data_format: ‘NHWC’ 和’NCHW’格式
• 返回一个tensor

Problem 1

原来基本的CNN模型中，卷积层步长stride为2，现在修改为使用步长stride为2和kernel size为2的max pooling。

batch_size = 16
patch_size = 5
depth = 16
num_hidden = 64

graph = tf.Graph()

with graph.as_default():

    # Input data.
    tf_train_dataset = tf.placeholder(
    tf.float32, shape=(batch_size, image_size, image_size, num_channels))
    tf_train_labels = tf.placeholder(tf.float32, shape=(batch_size, num_labels))
    tf_valid_dataset = tf.constant(valid_dataset)
    tf_test_dataset = tf.constant(test_dataset)

    # Variables.
    layer1_weights = tf.Variable(tf.truncated_normal(
      [patch_size, patch_size, num_channels, depth], stddev=0.1))
    layer1_biases = tf.Variable(tf.zeros([depth]))
    layer2_weights = tf.Variable(tf.truncated_normal(
      [patch_size, patch_size, depth, depth], stddev=0.1))
    layer2_biases = tf.Variable(tf.constant(1.0, shape=[depth]))
    layer3_weights = tf.Variable(tf.truncated_normal(
      [image_size // 4 * image_size // 4 * depth, num_hidden], stddev=0.1))
    layer3_biases = tf.Variable(tf.constant(1.0, shape=[num_hidden]))
    layer4_weights = tf.Variable(tf.truncated_normal(
      [num_hidden, num_labels], stddev=0.1))
    layer4_biases = tf.Variable(tf.constant(1.0, shape=[num_labels]))

    # Model with Max Pooling
    def model(data):
        conv = tf.nn.conv2d(data, layer1_weights, [1, 1, 1, 1], padding='SAME') # update stride as 1
        hidden = tf.nn.relu(conv + layer1_biases)
        pool = tf.nn.max_pool(hidden, [1, 2, 2, 1], [1, 2, 2, 1], padding='SAME') # add max pooling
        conv = tf.nn.conv2d(pool, layer2_weights, [1, 1, 1, 1], padding='SAME') # update stride as 1
        hidden = tf.nn.relu(conv + layer2_biases)
        pool = tf.nn.max_pool(hidden, [1, 2, 2, 1], [1, 2, 2, 1], padding='SAME') # add max pooling
        shape = pool.get_shape().as_list()
        reshape = tf.reshape(pool, [shape[0], shape[1] * shape[2] * shape[3]])
        hidden = tf.nn.relu(tf.matmul(reshape, layer3_weights) + layer3_biases)
        return tf.matmul(hidden, layer4_weights) + layer4_biases

    # Training computation.
    logits = model(tf_train_dataset)
    loss = tf.reduce_mean(
    tf.nn.softmax_cross_entropy_with_logits(labels=tf_train_labels, logits=logits))

    # Optimizer.
    optimizer = tf.train.GradientDescentOptimizer(0.05).minimize(loss)

    # Predictions for the training, validation, and test data.
    train_prediction = tf.nn.softmax(logits)
    valid_prediction = tf.nn.softmax(model(tf_valid_dataset))
    test_prediction = tf.nn.softmax(model(tf_test_dataset))

num_steps = 3001

with tf.Session(graph=graph) as session:
    tf.global_variables_initializer().run()
    print('Initialized')
    for step in range(num_steps):
        offset = (step * batch_size) % (train_labels.shape[0] - batch_size)
        batch_data = train_dataset[offset:(offset + batch_size), :, :, :]
        batch_labels = train_labels[offset:(offset + batch_size), :]
        feed_dict = {tf_train_dataset : batch_data, tf_train_labels : batch_labels}
        _, l, predictions = session.run(
          [optimizer, loss, train_prediction], feed_dict=feed_dict)
        if (step % 500 == 0):
            print('Minibatch loss at step %d: %f' % (step, l))
            print('Minibatch accuracy: %.1f%%' % accuracy(predictions, batch_labels))
            print('Validation accuracy: %.1f%%' % accuracy(
            valid_prediction.eval(), valid_labels))
    print('Test accuracy: %.1f%%' % accuracy(test_prediction.eval(), test_labels))

结果如下：

Initialized
Minibatch loss at step 0: 3.217397
Minibatch accuracy: 6.2%
Validation accuracy: 10.2%
Minibatch loss at step 500: 0.567970
Minibatch accuracy: 75.0%
Validation accuracy: 81.4%
Minibatch loss at step 1000: 0.269530
Minibatch accuracy: 87.5%
Validation accuracy: 84.0%
Minibatch loss at step 1500: 0.404251
Minibatch accuracy: 81.2%
Validation accuracy: 85.3%
Minibatch loss at step 2000: 0.741879
Minibatch accuracy: 81.2%
Validation accuracy: 85.9%
Minibatch loss at step 2500: 0.048788
Minibatch accuracy: 100.0%
Validation accuracy: 86.0%
Minibatch loss at step 3000: 0.787112
Minibatch accuracy: 75.0%
Validation accuracy: 86.7%
Test accuracy: 93.1%

Problem 2

呃，还没来得及写完……

TensorFlow的python API文档(https://www.tensorflow.org/api_docs/python/)是个好东西，常用的函数接口如tf.nn.conv2d、tf.nn.max_pool、tf.nn.relu等都有详细介绍。