Tensorflow学习笔记3：卷积神经网络实现手写字符识别

# -*- coding:utf-8 -*-
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
import os
import argparse
import sys

DATA_DIR = os.path.join('.', 'mnist_link')

# =======================================
#            COMMON OPERATIONS
# =======================================
def conv2d(x, W):
    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')

def max_pool_2x2(x):
    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')

def init_weight(shape):
    init = tf.truncated_normal(shape, stddev=0.1)
    return tf.Variable(init)

def init_bias(shape):
    init = tf.constant(0.1, shape=shape)
    return tf.Variable(init)

# =======================================
#              BUILD CNN
# =======================================
def build_cnn(x):
    '''
    build the cnn model
    '''
    x_image = tf.reshape(x, [-1,28,28,1])

    w1 = init_weight([5,5,1,32])
    b1=init_bias([32])
    conv1 = tf.nn.relu(conv2d(x_image, w1) + b1)
    pool1 = max_pool_2x2(conv1)

    w2 = init_weight([5,5,32,64])
    b2 = init_bias([64])
    conv2 = tf.nn.relu(conv2d(pool1, w2) + b2)
    pool2 = max_pool_2x2(conv2)

    # fc
    w_fc1 = init_weight([7*7*64, 1024])
    b_fc1 = init_bias([1024])
    pool2_flat = tf.reshape(pool2, [-1, 7*7*64])
    fc1 = tf.nn.relu(tf.matmul(pool2_flat, w_fc1) + b_fc1)

    # dropout
    keep_prob = tf.placeholder(tf.float32)
    fc1_dropout = tf.nn.dropout(fc1, keep_prob)

    # fc2
    w_fc2 = init_weight([1024, 10])
    b_fc2 = init_bias([10])
    y_conv = tf.matmul(fc1_dropout, w_fc2) + b_fc2
    return y_conv, keep_prob


# =======================================
#            train and test
# =======================================
def main():
    '''
    feed data into cnn model, and train and test the model
    '''
    # import data
    print('import data...')
    mnist = input_data.read_data_sets(DATA_DIR, one_hot=True)

    # create graph for cnn
    x = tf.placeholder(tf.float32, [None, 784])
    y_ = tf.placeholder(tf.float32, [None, 10])
    y_conv, keep_prob = build_cnn(x)

    cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels = y_, logits = y_conv))
    optimizer = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
    correct_predictions = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))
    accuracy = tf.reduce_mean(tf.cast(correct_predictions, tf.float32))
    init = tf.global_variables_initializer()

    print('start training...')
    with tf.Session() as sess:
        sess.run(init)
        for i in range(2000):
            batch = mnist.train.next_batch(128)
            optimizer.run(feed_dict={x:batch[0], y_:batch[1], keep_prob:0.5})
            if i%100 == 0:
                train_acc = accuracy.eval(
                    feed_dict = {x:batch[0], y_:batch[1], keep_prob:1.0})
                print('step {}, accuracy is {}'.format(i, train_acc))
        
        test_acc = accuracy.eval(feed_dict={x:mnist.test.images, y_:mnist.test.labels, keep_prob:1.0})
        print('test accuracy is {}'.format(test_acc))


if __name__ == '__main__':
    print('run main')
    main()