def restore_model(feedData):
#复现计算图
with tf.Graph().as_default() as g:
#给x占位
x = tf.placeholder(tf.float32, [None, face_forward.INPUT_NODE])
y = face_forward.forward(x, None)
preValue = tf.argmax(y, 1) #
variable_average = tf.train.ExponentialMovingAverage(
face_backward.MOVING_AVERAGE_DECAY)
variable_restore = variable_average.variables_to_restore()
saver = tf.train.Saver(variable_restore)
with tf.Session() as sess:
#加载CKPT
ckpt = tf.train.get_checkpoint_state(face_backward.MODEL_SAVE_PATH)
#判断是否有模型
if ckpt and ckpt.model_checkpoint_path:
#恢复模型到当前会话
saver.restore(sess, ckpt.model_checkpoint_path)
#执行预测操作
preValue = sess.run(preValue, feed_dict={x: feedData})
#观察
preArr = sess.run(y, feed_dict={x: feedData})
# preArr = preArr.reshape([1,10])
print(preArr)
return preValue
else:
print('No checkpoint file found')
return -1
def backward():
x = tf.placeholder(tf.float32, [None, face_forward.INPUT_NODE])
y_ = tf.placeholder(tf.float32, [None, face_forward.OUTPUT_NODE])
y = face_forward.forward(x, REGULARIZER)
global_step = tf.Variable(0, trainable=False)
ce = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=y,
labels=tf.argmax(
y_, 1))
cem = tf.reduce_mean(ce)
loss = cem + tf.add_n(tf.get_collection("losses"))
learning_rate = tf.train.exponential_decay(LEARNING_RATE_BASE,
global_step,
train_num_examples / BATCH_SIZE,
LEARNING_RATE_DECAY,
staircase=True)
train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(
loss, global_step=global_step)
ema = tf.train.ExponentialMovingAverage(MOVING_AVERAGE_DECAY, global_step)
ema_op = ema.apply(tf.trainable_variables())
with tf.control_dependencies([train_step, ema_op]):
train_op = tf.no_op(name="train")
saver = tf.train.Saver()
img_batch, label_batch = face_generateds.get_tfRecode(BATCH_SIZE,
isTrain=True)
with tf.Session() as sess:
init_op = tf.initialize_all_variables()
sess.run(init_op)
#duan dian ji xu
ckpt = tf.train.get_checkpoint_state(MODEL_SAVE_PATH)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess, ckpt.model_checkpoint_path)
#thread start
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(sess=sess, coord=coord)
for i in range(STEPS):
xs, ys = sess.run([img_batch, label_batch])
_, loss_value, step = sess.run([train_op, loss, global_step],
feed_dict={
x: xs,
y_: ys
})
if i % 100 == 0:
print("After %d training steps,loss on training batch is % g" %
(step, loss_value))
saver.save(sess,
os.path.join(MODEL_SAVE_PATH, MODEL_NAME),
global_step=global_step)
coord.request_stop()
coord.join(threads)
def test(face_data):
#复现计算图
with tf.Graph().as_default() as g:
x = tf.placeholder(tf.float32, [None, face_forward.INPUT_NODE])
y_ = tf.placeholder(tf.float32, [None, face_forward.OUTPUT_NODE])
y = face_forward.forward(x, None)
#恢复过后，数据会得到各自的滑动平均值
ema = tf.train.ExponentialMovingAverage(
face_backward.MOVING_AVERAGE_DECAY)
ema_restore = ema.variables_to_restore()
saver = tf.train.Saver(ema_restore)
correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
#拿数据
feature, label = face_generateds.get_tfrecord(TEST_NUM, isTrain=False)
while True:
with tf.Session() as sess:
#滑动平均值赋值给各个参数
ckpt = tf.train.get_checkpoint_state(
face_backward.MODEL_SAVE_PATH)
#判断是否有模型
if ckpt and ckpt.model_checkpoint_path:
#恢复模型到当前会话
saver.restore(sess, ckpt.model_checkpoint_path)
#恢复global_step值
global_step = ckpt.model_checkpoint_path.split(
'/')[-1].split('-')[-1]
#线程调节器
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(sess=sess,
coord=coord)
xs, ys = sess.run([feature, label])
#执行准确率计算
accuracy_score = sess.run(accuracy,
feed_dict={
x: xs,
y_: ys
})
#打印准确率
print("after %s training steps , test accuracy = %g" %
(global_step, accuracy_score))
else:
print('No checkpoint file found')
return
time.sleep(TEST_INTERVAL_SECS)
def restore_model(testPicArr):
with tf.Graph().as_default() as tg:
x = tf.placeholder(tf.float32, [None, face_forward.INPUT_NODE])
y = face_forward.forward(x, None)
preValue = tf.argmax(y, 1)
variable_averages = tf.train.ExponentialMovingAverage(
face_backword.MOVING_AVERAGE_DECAY)
variables_to_restore = variable_averages.variables_to_restore()
saver = tf.train.Saver(variables_to_restore)
with tf.Session() as sess:
ckpt = tf.train.get_checkpoint_state(face_backword.MODEL_SAVE_PATH)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess, ckpt.model_checkpoint_path)
preValue = sess.run(preValue, feed_dict={x: testPicArr})
return preValue
else:
print " No checkPoint file found"
return -1
def test():
with tf.Graph().as_default() as g:
x = tf.placeholder(tf.float32, [None, face_forward.INPUT_NODE])
y_ = tf.placeholder(tf.float32, [None, face_forward.OUTPUT_NODE])
y = face_forward.forward(x, None)
ema = tf.train.ExponentialMovingAverage(
face_backword.MOVING_AVERAGE_DECAY)
ema_restore = ema.variables_to_restore()
saver = tf.train.Saver(ema_restore)
correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
img_batch, label_batch = face_generateds.get_tfRecode(TEST_NUM,
isTrain=False)
while True:
with tf.Session() as sess:
ckpt = tf.train.get_checkpoint_state(
face_backword.MODEL_SAVE_PATH)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess, ckpt.model_checkpoint_path)
global_step = ckpt.model_checkpoint_path.split(
'/')[-1].split('-')[-1]
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(sess=sess,
coord=coord)
xs, ys = sess.run([img_batch, label_batch])
accuracy_score = sess.run(accuracy,
feed_dict={
x: xs,
y_: ys
})
print("After %s traing steps (s),test accuracy = % g" %
(global_step, accuracy_score))
else:
print("Without checkpoint")
time.sleep(TIME_INTERVAL_SECS)
def backward(face_data):
x = tf.placeholder(tf.float32,[None,face_forward.INPUT_NODE])
y_ = tf.placeholder(tf.float32,[None,face_forward.OUTPUT_NODE])
#调用前向传播的程序，输出y
y = face_forward.forward(x,REGULARIZER)
#轮数计数器
global_step = tf.Variable(0,trainable = False)
#交叉熵
ce = tf.nn.sparse_softmax_cross_entropy_with_logits(logits = y,labels = tf.argmax(y_,1))
cem = tf.reduce_mean(ce)
#定义损失函数
loss = cem + tf.add_n(tf.get_collection('losses'))
learning_rate = tf.train.exponential_decay(
LEARNING_RATE_BASE,
global_step,
train_num_examples/BATCH_SIZE,
LEARNING_RATE_DECAY,
staircase = True)
#定义训练过程：这个是梯度下降优化程序
train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss,global_step = global_step)
#定义滑动平均
ema = tf.train.ExponentialMovingAverage(MOVING_AVERAGE_DECAY,global_step)
#每一次运行这一句，所有待优化参数求滑动平均
ema_op = ema.apply(tf.trainable_variables())
with tf.control_dependencies([train_step,ema_op]):
train_op = tf.no_op(name = 'train')
saver = tf.train.Saver()
#自己写的特征和标签
feature,label = face_generateds.get_tfrecord(BATCH_SIZE,isTrain = True)
with tf.Session() as sess:
init_op = tf.global_variables_initializer()
sess.run(init_op)
#加载ckpt模型,实现断点续训 #给所有的w和b赋值保存再ckpt中的值,实现断点续训
ckpt = tf.train.get_checkpoint_state(MODEL_SAVE_PATH)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess,ckpt.model_checkpoint_path)
#加载完成
#开启线程调节器
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(sess = sess,coord = coord)
for i in range(STEPS):
#每次读入BATCH_SIZE组数据
xs,ys = sess.run([feature,label])
#喂入神经网络，执行训练过程
_,loss_value,step = sess.run([train_op,loss,global_step],feed_dict={x:xs,y_:ys})
if i % 1000 == 0:
print("after %d training steps,loss on training batch is %g" %(step,loss_value))
i = i % 1000
saver.save(sess,os.path.join(MODEL_SAVE_PATH,MODEL_NAME),global_step = global_step)
pass
#关闭线程调节器
coord.request_stop()
coord.join(threads)
pass