Q&A
ベストアンサー
回答ありがとうございます。
テストに関しましては、自分で用意した画像のデータを使用しております。
コードについてなのですが以下のようになっています。
お手数おかけしますが、よろしくお願いいたします。
import sys
import cv2
import math
import numpy as np
import tensorflow as tf
import tensorflow.python.platform
NUM_CLASSES = 10
IMAGE_SIZE = 224
IMAGE_PIXELS = IMAGE_SIZE*IMAGE_SIZE*3
flags = tf.app.flags
FLAGS = flags.FLAGS
flags.DEFINE_string('train', 'train.txt', 'File name of train data')
flags.DEFINE_string('test', 'test.txt', 'File name of train data')
flags.DEFINE_string('train_dir', '/tmp/data', 'Directory to put the training data.')
flags.DEFINE_integer('max_steps', 200, 'Number of steps to run trainer.')
flags.DEFINE_integer('batch_size', 100, 'Batch size'
'Must divide evenly into the dataset sizes.')
flags.DEFINE_float('learning_rate', 0.007, 'Initial learning rate.')
def inference(images_placeholder, keep_prob):
# 重みを標準偏差0.1の正規分布で初期化
def weight_variable(shape,num):
initial = tf.truncated_normal(shape, stddev=0.1/math.sqrt(float(num)))
# tf.Variable:計算グラフ上の変数シンボルinitialを定義
return (tf.Variable(initial).initialized_value())
def bias_variable(shape):
# 定数のtensorを作る
initial = tf.constant(0.1, shape=shape)
return (tf.Variable(initial).initialized_value())
# 畳み込み層の作成
def conv2d_first(x, W):
return (tf.nn.conv2d(x, W, strides=[1, 4, 4, 1], padding='SAME'))
def conv2d(x,W):
return (tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME'))
# プーリング層の作成
def max_pool_3x3(x):
return tf.nn.max_pool(x, ksize=[1, 3, 3, 1],
strides=[1, 2, 2, 1], padding='SAME')
# 入力を224x224x3に変形
x_image = tf.reshape(images_placeholder, [-1, IMAGE_SIZE, IMAGE_SIZE, 3])
# 畳み込み層1の作成
with tf.name_scope('conv1') as scope:
W_conv1 = weight_variable([11, 11, 3, 96], IMAGE_SIZE*IMAGE_SIZE)
b_conv1 = bias_variable([96])
h_conv1 = tf.nn.relu(conv2d_first(x_image, W_conv1 + b_conv1))
# プーリング層1の作成
with tf.name_scope('pool1') as scope:
h_pool1 = max_pool_3x3(tf.nn.local_response_normalization(h_conv1))
# 畳み込み層2の作成
with tf.name_scope('conv2') as scope:
W_conv2 = weight_variable([5, 5, 96, 256],96)
b_conv2 = bias_variable([256])
h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
# プーリング層2の作成
with tf.name_scope('pool2') as scope:
h_pool2 = max_pool_3x3(tf.nn.local_response_normalization(h_conv2))
# 畳み込み層3の作成
with tf.name_scope('conv3') as scope:
W_conv3 = weight_variable([3,3,256,384], 256)
b_conv3 = bias_variable([384])
h_conv3 = tf.nn.relu(conv2d(h_pool2, W_conv3) + b_conv3)
# 畳み込み層4の作成
with tf.name_scope('conv4') as scope:
W_conv4 = weight_variable([3,3,384,384], 384)
b_conv4 = bias_variable([384])
h_conv4 = tf.nn.relu(conv2d(h_conv3, W_conv4) + b_conv4)
# 畳み込み層5の作成
with tf.name_scope('conv5') as scope:
W_conv5 = weight_variable([3,3,384,256], 384)
b_conv5 = bias_variable([256])
h_conv5 = tf.nn.relu(conv2d(h_conv4, W_conv5) + b_conv5)
# プーリング層3の作成
with tf.name_scope('pool3') as scope:
h_pool3 = max_pool_3x3(h_conv5)
# 全結合層1の作成
with tf.name_scope('fc1') as scope:
W_fc1 = weight_variable([7*7*256, 4096], (7*7*256))
b_fc1 = bias_variable([4096])
h_pool3_flat = tf.reshape(h_pool3, [-1, 7*7*256])
h_fc1 = tf.nn.relu(tf.matmul(h_pool3_flat, W_fc1) + b_fc1)
# dropoutの設定
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)
# 全結合層2の作成
with tf.name_scope('fc2') as scope:
w_fc2 = weight_variable([4096,4096], 4096)
b_fc2 = bias_variable([4096])
h_fc2 = tf.nn.relu(tf.matmul(h_fc1_drop,w_fc2) + b_fc2)
h_fc2_drop = tf.nn.dropout(h_fc2, keep_prob)
# 全結合層3の作成
with tf.name_scope('fc3') as scope:
W_fc3 = weight_variable([4096, NUM_CLASSES], 4096)
b_fc3 = bias_variable([NUM_CLASSES])
y_conv = tf.matmul(h_fc2_drop, W_fc3) + b_fc3
# ソフトマックス関数による正規化
with tf.name_scope('softmax') as scope:
y_conv=tf.nn.softmax(tf.matmul(h_fc2_drop, W_fc3) + b_fc3)
# 各ラベルの確率のようなものを返す
return y_conv
def loss(logits, labels):
""" lossを計算する関数
引数:
logits: ロジットのtensor, float - [batch_size, NUM_CLASSES]
labels: ラベルのtensor, int32 - [batch_size, NUM_CLASSES]
返り値:
cross_entropy: 交差エントロピーのtensor, float
"""
# 交差エントロピーの計算
cross_entropy = tf.reduce_mean(-tf.reduce_sum(labels*tf.log(tf.clip_by_value(logits,1e-10,1.0))))
# TensorBoardで表示するよう指定
tf.summary.scalar("cross_entropy", cross_entropy)
return cross_entropy
def training(loss, learning_rate):
train_step = tf.train.AdamOptimizer(learning_rate).minimize(loss)
return train_step
def accuracy(logits, labels):
correct_prediction = tf.equal(tf.argmax(logits, 1), tf.argmax(labels, 1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
tf.summary.scalar("accuracy", accuracy)
return accuracy
if __name__ == '__main__':
# ファイルを開く
f = open(FLAGS.train, 'r')
# データを入れる配列
train_image = []
train_label = []
for line in f:
# 改行を除いてスペース区切りにする
line = line.rstrip()
l = line.split()
# # データを読み込んでIMAGE_SIZExIMAGE_SIZEに縮小
img = cv2.imread(l[0])
img = cv2.resize(img, (IMAGE_SIZE, IMAGE_SIZE))
# 一列にした後、0-1のfloat値にする
train_image.append(img.flatten().astype(np.float32)/255.0)
# ラベルを1-of-k方式で用意する
tmp = np.zeros(NUM_CLASSES)
tmp[int(l[1])] = 1
train_label.append(tmp)
# numpy形式に変換
train_image = np.asarray(train_image)
train_label = np.asarray(train_label)
f.close()
f = open(FLAGS.test, 'r')
test_image = []
test_label = []
for line in f:
line = line.rstrip()
l = line.split()
img = cv2.imread(l[0])
img = cv2.resize(img, (IMAGE_SIZE, IMAGE_SIZE))
test_image.append(img.flatten().astype(np.float32)/255.0)
tmp = np.zeros(NUM_CLASSES)
tmp[int(l[1])] = 1
test_label.append(tmp)
test_image = np.asarray(test_image)
test_label = np.asarray(test_label)
f.close()
with tf.Graph().as_default():
# 画像を入れる仮のTensor
images_placeholder = tf.placeholder("float32", shape=(None, IMAGE_PIXELS))
# ラベルを入れる仮のTensor
labels_placeholder = tf.placeholder("float32", shape=(None, NUM_CLASSES))
# dropout率を入れる仮のTensor
keep_prob = tf.placeholder("float")
# inference()を呼び出してモデルを作る
logits = inference(images_placeholder, keep_prob)
# loss()を呼び出して損失を計算
loss_value = loss(logits, labels_placeholder)
# training()を呼び出して訓練
train_op = training(loss_value, FLAGS.learning_rate)
# 精度の計算
acc = accuracy(logits, labels_placeholder)
# 保存の準備
saver = tf.train.Saver()
# Sessionの作成
sess = tf.Session()
# 変数の初期化
sess.run(tf.global_variables_initializer())
# TensorBoardで表示する値の設定
summary_op = tf.summary.merge_all()
summary_writer = tf.summary.FileWriter(FLAGS.train_dir, sess.graph_def)
# 訓練の実行
for step in range(FLAGS.max_steps):
for i in range(len(train_image)//FLAGS.batch_size):
# batch_size分の画像に対して訓練の実行
batch = FLAGS.batch_size*i
# feed_dictでplaceholderに入れるデータを指定する
point=sess.run(train_op, feed_dict={
images_placeholder: train_image[batch:batch+FLAGS.batch_size],
labels_placeholder: train_label[batch:batch+FLAGS.batch_size],
keep_prob: 0.5})
# 1 step終わるたびに精度を計算する
train_accuracy = sess.run(acc, feed_dict={
images_placeholder: train_image,
labels_placeholder: train_label,
keep_prob: 1.0})
print ("step %d, training accuracy %g"%(step, train_accuracy))
# 1 step終わるたびにTensorBoardに表示する値を追加する
summary_str = sess.run(summary_op, feed_dict={
images_placeholder: train_image,
labels_placeholder: train_label,
keep_prob: 1.0})
summary_writer.add_summary(summary_str, step)
# 訓練が終了したらテストデータに対する精度を表示
print ("test accuracy %g"%sess.run(acc, feed_dict={
images_placeholder: test_image,
labels_placeholder: test_label,
keep_prob: 1.0}))
# 最終的なモデルを保存
save_path = saver.save(sess, "model.ckpt")
# 画像を入れる仮のTensor
images_placeholder = tf.placeholder("float32", shape=(None, IMAGE_PIXELS))
# ラベルを入れる仮のTensor
labels_placeholder = tf.placeholder("float32", shape=(None, NUM_CLASSES))
# dropout率を入れる仮のTensor
keep_prob = tf.placeholder("float32")
logits = inference(images_placeholder,keep_prob)
sess = tf.InteractiveSession()
saver = tf.train.Saver()
hoge = sess.run(tf.global_variables_initializer())
saver.restore(sess,"model.ckpt")
for i in range(len(test_image)):
#最大の確率のラベルを表示
hoge = np.argmax(logits.eval(feed_dict={images_placeholder: [test_image[i]],keep_prob: 1.0 })[0])
print("%s"%hoge)
exit()
バッドをするには、ログインかつ
こちらの条件を満たす必要があります。
2017/11/09 03:41
2017/11/09 10:44
2017/11/09 13:32
2017/11/10 07:15