人工智能从入门到精通（15）-卷积网络在数字识别的应用

经典卷积网络模型

LeNet-5模型

1. 卷积层
2. 池化层
3. 卷积层
4. 池化层
5. 全连接层
6. 全连接层

7. 全连接层（近似）

   代码
   train

import os
import numpy as np
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

import mnist_inference



#data params
BATCH_SIZE = 100
LEARNING_RATE_BASE = 0.8
LEARNING_RATE_DECAY = 0.99
REGULARAZTION_RATE = 0.0001
TRAINING_STEPS = 20000
MOVING_AVERAGE_DECAY = 0.99

#model save path and name
MODEL_SAVE_PATH="/path/to/model"
MODEL_NAME = "model.ckpt"

def train(mnist):
    x = tf.placeholder(tf.float32,[BATCH_SIZE,mnist_inference.IMAGE_SIZE,
                                   mnist_inference.IMAGE_SIZE,mnist_inference.NUM_CHANNELS],name='x-input')
    y_ = tf.placeholder(tf.float32,[None,mnist_inference.OUTPUT_NODE],name='y-input')


    regularizer = tf.contrib.layers.l2_regularizer(REGULARAZTION_RATE)

    y = mnist_inference.inference(x,1,regularizer)

    #step to control the delay
    global_step = tf.Variable(0,trainable=False)

    variable_averages = tf.train.ExponentialMovingAverage(MOVING_AVERAGE_DECAY,global_step)
    variable_averages_op = variable_averages.apply(tf.trainable_variables())

    #cross entropy and add the regularization
    # cross_entropy = -tf.reduce_sum(y_*tf.log(y))
    cross_entropy = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=y,labels=tf.argmax(y_,1))
    cross_entropy_mean = tf.reduce_mean(cross_entropy)
    loss = cross_entropy_mean + tf.add_n(tf.get_collection('losses'))

    learning_rate = tf.train.exponential_decay(LEARNING_RATE_BASE,global_step,
                                               mnist.train.num_examples/BATCH_SIZE,
                                               LEARNING_RATE_DECAY)

    # tf.scalar_summary('learning_rate', learning_rate)
    train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss,global_step=global_step)


    #updata the W and variable average at the same time
    with tf.control_dependencies([train_step,variable_averages_op]):
        train_op = tf.no_op(name='train')

    #save the model
    saver = tf.train.Saver()
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())

        for i in range(TRAINING_STEPS):
            xs,ys = mnist.train.next_batch(BATCH_SIZE)
            reshaped_xs = np.reshape(xs,(BATCH_SIZE,mnist_inference.IMAGE_SIZE,
                                         mnist_inference.IMAGE_SIZE,mnist_inference.NUM_CHANNELS))
            _, loss_value,step = sess.run([train_op, loss,global_step], feed_dict={x: reshaped_xs, y_: ys})
            if i % 1000 == 0:
                # print "step %d, training accuracy %g" % (i, train_accuracy)
                print ("step %d,loss is  %g" % (step,loss_value))
            saver.save(sess,os.path.join(MODEL_SAVE_PATH,MODEL_NAME),global_step=global_step)

def main(argv=None):
    mnist = input_data.read_data_sets("/path/to/MNIST_data/", one_hot=True)
    train(mnist)

if __name__=='__main__':
    tf.app.run()

inference

import tensorflow as tf
import numpy as np
from tensorflow.examples.tutorials.mnist import input_data

#dataset params
INPUT_NODE = 784
OUTPUT_NODE = 10

#cnns params
IMAGE_SIZE = 28
NUM_CHANNELS =1
NUM_LABELS =10

DROP_PROB = 0.5

CON1_DEEP = 32
CON1_SIZE = 5

CON2_DEEP = 64
CON2_SIZE = 5

FC1_SIZE = 1024

#inference structure
def inference(input_tensor,train,regularizer):
    #first layer
    with tf.variable_scope('layer1-conv1'):
        conv1_W = tf.get_variable("weight",[CON1_SIZE,CON1_SIZE,NUM_CHANNELS,CON1_DEEP]
                                  ,initializer=tf.truncated_normal_initializer(stddev=0.1))
        conv1_b = tf.get_variable("bias",[CON1_DEEP],initializer=tf.constant_initializer(0.1))

        #5 * 5 patch ,step 1 ,fill 0
        conv1 = tf.nn.conv2d(input_tensor,conv1_W,strides=[1,1,1,1],padding='SAME')
        relu1 = tf.nn.relu(tf.nn.bias_add(conv1,conv1_b))

    with tf.variable_scope('layer1-max_pool'):
        pool1 = tf.nn.max_pool(relu1,ksize=[1,2,2,1],strides=[1,2,2,1],padding='SAME')

    #second layer
    with tf.variable_scope('layer2-conv2'):
        conv2_W = tf.get_variable("weight",[CON2_SIZE,CON2_SIZE,CON1_DEEP,CON2_DEEP],
                                  initializer=tf.truncated_normal_initializer(stddev=0.1))
        conv2_b = tf.get_variable("bias",[CON2_DEEP],initializer=tf.constant_initializer(0.1))

        #5*5 patch,step 1 ,fill 0
        conv2 = tf.nn.conv2d(pool1,conv2_W,strides=[1,1,1,1],padding='SAME')
        relu2 = tf.nn.relu(tf.nn.bias_add(conv2,conv2_b))

    with tf.variable_scope('layer2-max_pool'):
        pool2 = tf.nn.max_pool(relu2,ksize=[1,2,2,1],strides=[1,2,2,1],padding='SAME')

    #fc1 layer
    pool_shape = pool2.get_shape().as_list()
    nodes = pool_shape[1] * pool_shape[2] * pool_shape[3]
    reshaped = tf.reshape(pool2,[-1,nodes])
    with tf.variable_scope('layer3-fc1'):
        fc1_W = tf.get_variable("weight",[nodes,FC1_SIZE],
                                initializer=tf.truncated_normal_initializer(stddev=0.1))
        #regularizer
        if regularizer != None:
            tf.add_to_collection('losses',regularizer(fc1_W))
        fc1_b = tf.get_variable("bias",[FC1_SIZE],initializer=tf.constant_initializer(0.1))
        fc1 = tf.nn.relu(tf.matmul(reshaped,fc1_W)+fc1_b)
        if train:
            fc1 = tf.nn.dropout(fc1,0.5)

    with tf.variable_scope('layer3-softmax'):
        fc2_W = tf.get_variable("weight",[FC1_SIZE,NUM_LABELS],
                                initializer=tf.truncated_normal_initializer(stddev=0.1))
        if regularizer != None:
            tf.add_to_collection('losses',regularizer(fc2_W))
        fc2_b = tf.get_variable("bias",[NUM_LABELS],initializer=tf.constant_initializer(0.1))
        y_conv = tf.nn.softmax(tf.matmul(fc1,fc2_W)+fc2_b)

    return y_conv

eval

import time
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
import numpy as np
import mnist_inference
import mnist_train

EVAL_INTERVAL_SECS = 10

def evaluate(mnist):
    with tf.Graph().as_default() as g:
        x = tf.placeholder(tf.float32, [None, mnist_inference.IMAGE_SIZE,
                                        mnist_inference.IMAGE_SIZE, mnist_inference.NUM_CHANNELS], name='x-input')
        y_ = tf.placeholder(tf.float32,[None,mnist_inference.OUTPUT_NODE],
                            name='y-input')
        xs = mnist.test.images
        reshaped_xs = np.reshape(xs, (-1, mnist_inference.IMAGE_SIZE,
                                      mnist_inference.IMAGE_SIZE, mnist_inference.NUM_CHANNELS))
        test_feed = {x:reshaped_xs,y_:mnist.test.labels}

        y = mnist_inference.inference(x,None,None)

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

        variable_averages = tf.train.ExponentialMovingAverage(mnist_train.MOVING_AVERAGE_DECAY)
        variable_to_restore = variable_averages.variables_to_restore()
        saver = tf.train.Saver(variable_to_restore)

        while True:
            with tf.Session() as sess:
                ckpt = tf.train.get_checkpoint_state(mnist_train.MODEL_SAVE_PATH)
                if ckpt and ckpt.model_checkpoint_path:
                    saver.restore(sess,ckpt.model_checkpoint_path)

                    global_stop = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]

                    accuracy_score = sess.run(accuracy,feed_dict=test_feed)
                    print ("step %s ,accuracy is %g" %(global_stop,accuracy_score))

                else:
                    print ("NOT FOUND FILE")
                    return
                time.sleep(EVAL_INTERVAL_SECS)

def main(argv=None):
    mnist = input_data.read_data_sets("/path/to/MNIST_data/", one_hot=True)
    evaluate(mnist)

if __name__=='__main__':
    tf.app.run()