2
4

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?

More than 5 years have passed since last update.

畳み込みニューラルネット Part5 [TensorFlowでDeep Learning 8]

Last updated at Posted at 2016-05-28

(目次はこちら)
(Tensorflow 2.0版: https://qiita.com/kumonkumon/items/bd009ae158f158b1a81b)

#はじめに
前回の記事では、フィルタを固定した畳み込みニューラルネットワークにドロップアウト層を追加した。今回は、Prewitte filterで固定していた畳み込み層のウェイトも学習によって推定するようにしてみる。

#畳み込み層のパラメータ推定
フィルタを固定した畳み込みニューラルネットワークを、フィルタについても推定する(学習する)ように拡張するだけ。

mnist_cnn_sl

##コード
Python: 3.6.8, Tensorflow: 1.13.1で動作確認済み
(もともと2016年前半に書いたものなので、順次更新しています。)

mnist_cnn_sl.py

mnist_cnn_sl.py
from helper import *

IMAGE_WIDTH, IMAGE_HEIGHT = 28, 28
CATEGORY_NUM = 10
LEARNING_RATE = 0.1
FILTER_SIZE = 5
FILTER_NUM = 32
FEATURE_DIM = 100
KEEP_PROB = 0.5
TRAINING_LOOP = 20000
BATCH_SIZE = 100
SUMMARY_DIR = 'log_cnn_sl'
SUMMARY_INTERVAL = 1000
BUFFER_SIZE = 1000
EPS = 1e-10


with tf.Graph().as_default():
    (X_train, y_train), (X_test, y_test) = mnist_samples()
    ds = tf.data.Dataset.from_tensor_slices((X_train, y_train))
    ds = ds.shuffle(BUFFER_SIZE).batch(BATCH_SIZE).repeat(int(TRAINING_LOOP * BATCH_SIZE / X_train.shape[0]) + 1)
    next_batch = ds.make_one_shot_iterator().get_next()

    with tf.name_scope('input'):
        y_ = tf.placeholder(tf.float32, [None, CATEGORY_NUM], name='labels')
        x = tf.placeholder(tf.float32, [None, IMAGE_HEIGHT, IMAGE_WIDTH], name='input_images')

    with tf.name_scope('convolution'):
        W_conv = weight_variable([FILTER_SIZE, FILTER_SIZE, 1, FILTER_NUM], name='weight_conv')
        b_conv = bias_variable([FILTER_NUM], name='bias_conv')
        x_image = tf.reshape(x, [-1, IMAGE_WIDTH, IMAGE_HEIGHT, 1])
        h_conv = tf.nn.relu(tf.nn.conv2d(x_image, W_conv, strides=[1, 1, 1, 1], padding='SAME') + b_conv)

    with tf.name_scope('pooling'):
        h_pool = tf.nn.max_pool(h_conv, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')

    with tf.name_scope('fully-connected'):
        W_fc = weight_variable([int(IMAGE_HEIGHT / 2 * IMAGE_WIDTH / 2 * FILTER_NUM), FEATURE_DIM], name='weight_fc')
        b_fc = bias_variable([FEATURE_DIM], name='bias_fc')
        h_pool_flat = tf.reshape(h_pool, [-1, int(IMAGE_HEIGHT / 2 * IMAGE_WIDTH / 2 * FILTER_NUM)])
        h_fc = tf.nn.relu(tf.matmul(h_pool_flat, W_fc) + b_fc)

    with tf.name_scope('dropout'):
        keep_prob = tf.placeholder(tf.float32)
        h_drop = tf.nn.dropout(h_fc, keep_prob)

    with tf.name_scope('readout'):
        W = weight_variable([FEATURE_DIM, CATEGORY_NUM], name='weight')
        b = bias_variable([CATEGORY_NUM], name='bias')
        y = tf.nn.softmax(tf.matmul(h_drop, W) + b)

    with tf.name_scope('optimize'):
        y = tf.clip_by_value(y, EPS, 1.0)
        cross_entropy = tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(y), axis=1))
        train_step = tf.train.GradientDescentOptimizer(LEARNING_RATE).minimize(cross_entropy)
        cross_entropy_summary = tf.summary.scalar('cross entropy', cross_entropy)

    with tf.Session() as sess:
        train_writer = tf.summary.FileWriter(SUMMARY_DIR + '/train', sess.graph)
        test_writer = tf.summary.FileWriter(SUMMARY_DIR + '/test')

        correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))
        accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
        accuracy_summary = tf.summary.scalar('accuracy', accuracy)

        sess.run(tf.global_variables_initializer())
        for i in range(TRAINING_LOOP + 1):
            images, labels = sess.run(next_batch)
            sess.run(train_step, {x: images, y_: labels, keep_prob: KEEP_PROB})

            if i % SUMMARY_INTERVAL == 0:
                train_acc, summary = sess.run(
                        [accuracy, tf.summary.merge([cross_entropy_summary, accuracy_summary])],
                        {x: images, y_: labels, keep_prob: 1.0})
                train_writer.add_summary(summary, i)
                test_acc, summary = sess.run(
                        [accuracy, tf.summary.merge([cross_entropy_summary, accuracy_summary])],
                        {x: X_test, y_: y_test, keep_prob: 1.0})
                test_writer.add_summary(summary, i)
                print(f'step: {i}, train-acc: {train_acc}, test-acc: {test_acc}')

##コードの説明
変更点を。

###畳み込み層
変更はここだけ。Prewittフィルタが学習対象の層になっている。

with tf.name_scope('convolution'):
    W_conv = weight_variable([FILTER_SIZE, FILTER_SIZE, 1, FILTER_NUM], name='weight_conv')
    b_conv = bias_variable([FILTER_NUM], name='bias_conv')
    x_image = tf.reshape(x, [-1, IMAGE_WIDTH, IMAGE_HEIGHT, 1])
    h_conv = tf.nn.relu(tf.nn.conv2d(x_image, W_conv, strides=[1, 1, 1, 1], padding='SAME'))

##結果
テストデータ(青線)での識別率は、99.1%程度。なんと99%の大台に。

image.png

#あとがき
今回は、固定していた畳み込み層のウェイトも学習によって推定するようにしてみました。次回の記事では、よりディープラーニングらしくしてみます。といっても、TensorFlowのTutorialのDeep MNIST for Expertsほぼそのものなんですけど。
(追記: Deep MNIST For Expertsは2019年時点ではもうありません)

2
4
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
2
4

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?