def testModelVariables(self):
batch_size = 5
height, width = 224, 224
num_classes = 1000
with self.test_session():
inputs = tf.random_uniform((batch_size, height, width, 3))
alexnet.alexnet_v2(inputs, num_classes)
expected_names = ['alexnet_v2/conv1/weights',
'alexnet_v2/conv1/biases',
'alexnet_v2/conv2/weights',
'alexnet_v2/conv2/biases',
'alexnet_v2/conv3/weights',
'alexnet_v2/conv3/biases',
'alexnet_v2/conv4/weights',
'alexnet_v2/conv4/biases',
'alexnet_v2/conv5/weights',
'alexnet_v2/conv5/biases',
'alexnet_v2/fc6/weights',
'alexnet_v2/fc6/biases',
'alexnet_v2/fc7/weights',
'alexnet_v2/fc7/biases',
'alexnet_v2/fc8/weights',
'alexnet_v2/fc8/biases',
]
model_variables = [v.op.name for v in slim.get_model_variables()]
self.assertSetEqual(set(model_variables), set(expected_names))
def testModelVariables(self):
batch_size = 5
height, width = 224, 224
num_classes = 1000
with self.test_session():
inputs = tf.random.uniform((batch_size, height, width, 3))
alexnet.alexnet_v2(inputs, num_classes)
expected_names = ['alexnet_v2/conv1/weights',
'alexnet_v2/conv1/biases',
'alexnet_v2/conv2/weights',
'alexnet_v2/conv2/biases',
'alexnet_v2/conv3/weights',
'alexnet_v2/conv3/biases',
'alexnet_v2/conv4/weights',
'alexnet_v2/conv4/biases',
'alexnet_v2/conv5/weights',
'alexnet_v2/conv5/biases',
'alexnet_v2/fc6/weights',
'alexnet_v2/fc6/biases',
'alexnet_v2/fc7/weights',
'alexnet_v2/fc7/biases',
'alexnet_v2/fc8/weights',
'alexnet_v2/fc8/biases',
]
model_variables = [v.op.name for v in slim.get_model_variables()]
self.assertSetEqual(set(model_variables), set(expected_names))
def testForward(self):
batch_size = 1
height, width = 224, 224
with self.test_session() as sess:
inputs = tf.random_uniform((batch_size, height, width, 3))
logits, _ = alexnet.alexnet_v2(inputs)
sess.run(tf.global_variables_initializer())
output = sess.run(logits)
self.assertTrue(output.any())
def testForward(self):
batch_size = 1
height, width = 224, 224
with self.test_session() as sess:
inputs = tf.random.uniform((batch_size, height, width, 3))
logits, _ = alexnet.alexnet_v2(inputs)
sess.run(tf.compat.v1.global_variables_initializer())
output = sess.run(logits)
self.assertTrue(output.any())
def choose_model(x, model):
"""
选择模型
:param x:
:param model:
:return:
"""
# 模型保存路径，模型名，预训练文件路径，前向传播
if model == 'Alex':
log_dir = "E:/alum/log/Alex"
y, _ = alexnet.alexnet_v2(
x,
num_classes=CLASSES, # 分类的类别
is_training=True, # 是否在训练
dropout_keep_prob=1.0, # 保留比率
spatial_squeeze=True, # 压缩掉1维的维度
global_pool=GLOBAL_POOL) # 输入不是规定的尺寸时，需要global_pool
elif model == 'VGG':
log_dir = "E:/alum/log/VGG"
y, _ = vgg.vgg_16(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
spatial_squeeze=True,
global_pool=GLOBAL_POOL)
elif model == 'VGG2':
log_dir = "E:/alum/log/VGG2"
y, _ = vgg.vgg_16(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
spatial_squeeze=True,
global_pool=GLOBAL_POOL)
elif model == 'Incep4':
log_dir = "E:/alum/log/Incep4"
y, _ = inception_v4.inception_v4(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
reuse=None,
scope='InceptionV4',
create_aux_logits=True)
elif model == 'Res':
log_dir = "E:/alum/log/Res"
y, _ = resnet_v2.resnet_v2_50(x,
num_classes=CLASSES,
is_training=True,
global_pool=GLOBAL_POOL,
output_stride=None,
spatial_squeeze=True,
reuse=None,
scope='resnet_v2_50')
else:
print('Error: model name not exist')
return
return y, log_dir
def testFullyConvolutional(self):
batch_size = 1
height, width = 300, 400
num_classes = 1000
with self.test_session():
inputs = tf.random.uniform((batch_size, height, width, 3))
logits, _ = alexnet.alexnet_v2(inputs, num_classes, spatial_squeeze=False)
self.assertEquals(logits.op.name, 'alexnet_v2/fc8/BiasAdd')
self.assertListEqual(logits.get_shape().as_list(),
[batch_size, 4, 7, num_classes])
def testBuild(self):
batch_size = 5
height, width = 224, 224
num_classes = 1000
with self.test_session():
inputs = tf.random_uniform((batch_size, height, width, 3))
logits, _ = alexnet.alexnet_v2(inputs, num_classes)
self.assertEquals(logits.op.name, 'alexnet_v2/fc8/squeezed')
self.assertListEqual(logits.get_shape().as_list(),
[batch_size, num_classes])
def testFullyConvolutional(self):
batch_size = 1
height, width = 300, 400
num_classes = 1000
with self.test_session():
inputs = tf.random_uniform((batch_size, height, width, 3))
logits, _ = alexnet.alexnet_v2(inputs, num_classes, spatial_squeeze=False)
self.assertEquals(logits.op.name, 'alexnet_v2/fc8/BiasAdd')
self.assertListEqual(logits.get_shape().as_list(),
[batch_size, 4, 7, num_classes])
def testBuild(self):
batch_size = 5
height, width = 224, 224
num_classes = 1000
with self.test_session():
inputs = tf.random.uniform((batch_size, height, width, 3))
logits, _ = alexnet.alexnet_v2(inputs, num_classes)
self.assertEquals(logits.op.name, 'alexnet_v2/fc8/squeezed')
self.assertListEqual(logits.get_shape().as_list(),
[batch_size, num_classes])
def predict(root_path, model='Alex'):
# 占位符
x = tf.placeholder(tf.float32, [None, IMG_SIZE, IMG_SIZE, 1])
# 模型保存路径，前向传播
if model == 'Alex':
log_path = "../log/Alex"
y, _ = alexnet.alexnet_v2(
x,
num_classes=CLASSES, # 分类的类别
is_training=True, # 是否在训练
dropout_keep_prob=1.0, # 保留比率
spatial_squeeze=True, # 压缩掉1维的维度
global_pool=GLOBAL_POOL) # 输入不是规定的尺寸时，需要global_pool
else:
log_path = '../log/My'
y, _ = lenet5(x)
saver = tf.train.Saver()
with tf.Session() as sess:
# 恢复模型权重
print('Reading checkpoints: ', model)
ckpt = tf.train.get_checkpoint_state(log_path)
if ckpt and ckpt.model_checkpoint_path:
global_step = ckpt.model_checkpoint_path.split('/')[-1].split(
'-')[-1]
saver.restore(sess, ckpt.model_checkpoint_path)
print('Loading success, global_step is %s' % global_step)
else:
print('Error: no checkpoint file found')
return
file_list = os.listdir(root_path)
for file_name in file_list:
if file_name.split('.')[-1] == 'jpg':
img = Image.open(os.path.join(root_path, file_name))
img_show = img.resize((SHOW_SIZE, SHOW_SIZE))
img_show = np.array(img_show, np.uint8)
img = img.resize((IMG_SIZE, IMG_SIZE))
img = np.array(img, np.uint8)
img = np.expand_dims(img, axis=0)
img = np.expand_dims(img, axis=3)
pre = np.squeeze(sess.run(y, feed_dict={x: img}))
print('predict:', pre)
points = [[pre[0] * SHOW_SIZE, pre[1] * SHOW_SIZE],
[pre[2] * SHOW_SIZE, pre[3] * SHOW_SIZE],
[pre[4] * SHOW_SIZE, pre[5] * SHOW_SIZE],
[pre[6] * SHOW_SIZE, pre[7] * SHOW_SIZE]]
points = np.int0(points)
cv2.polylines(img_show, [points], True, (0, 0, 255), 1)
cv2.imshow('img_show', img_show)
cv2.waitKey()
cv2.destroyAllWindows()
def testGlobalPool(self):
batch_size = 1
height, width = 256, 256
num_classes = 1000
with self.test_session():
inputs = tf.random.uniform((batch_size, height, width, 3))
logits, _ = alexnet.alexnet_v2(inputs, num_classes, spatial_squeeze=False,
global_pool=True)
self.assertEquals(logits.op.name, 'alexnet_v2/fc8/BiasAdd')
self.assertListEqual(logits.get_shape().as_list(),
[batch_size, 1, 1, num_classes])
def testEvaluation(self):
batch_size = 2
height, width = 224, 224
num_classes = 1000
with self.test_session():
eval_inputs = tf.random_uniform((batch_size, height, width, 3))
logits, _ = alexnet.alexnet_v2(eval_inputs, is_training=False)
self.assertListEqual(logits.get_shape().as_list(),
[batch_size, num_classes])
predictions = tf.argmax(logits, 1)
self.assertListEqual(predictions.get_shape().as_list(), [batch_size])
def testEvaluation(self):
batch_size = 2
height, width = 224, 224
num_classes = 1000
with self.test_session():
eval_inputs = tf.random.uniform((batch_size, height, width, 3))
logits, _ = alexnet.alexnet_v2(eval_inputs, is_training=False)
self.assertListEqual(logits.get_shape().as_list(),
[batch_size, num_classes])
predictions = tf.argmax(input=logits, axis=1)
self.assertListEqual(predictions.get_shape().as_list(), [batch_size])
def Alex_net(image, reuse=tf.AUTO_REUSE, keep_prop=0.5):
image = tf.reshape(image, [-1, 224, 224, 3])
with tf.variable_scope(name_or_scope='Alex', reuse=reuse):
arg_scope = alexnet.alexnet_v2_arg_scope()
with slim.arg_scope(arg_scope):
logits, end_point = alexnet.alexnet_v2(image,
1000,
is_training=True,
dropout_keep_prob=keep_prop)
probs = tf.nn.softmax(logits) # probabilities
return logits, probs, end_point
def testTrainEvalWithReuse(self):
train_batch_size = 2
eval_batch_size = 1
train_height, train_width = 224, 224
eval_height, eval_width = 300, 400
num_classes = 1000
with self.test_session():
train_inputs = tf.random_uniform(
(train_batch_size, train_height, train_width, 3))
logits, _ = alexnet.alexnet_v2(train_inputs)
self.assertListEqual(logits.get_shape().as_list(),
[train_batch_size, num_classes])
tf.get_variable_scope().reuse_variables()
eval_inputs = tf.random_uniform(
(eval_batch_size, eval_height, eval_width, 3))
logits, _ = alexnet.alexnet_v2(eval_inputs, is_training=False,
spatial_squeeze=False)
self.assertListEqual(logits.get_shape().as_list(),
[eval_batch_size, 4, 7, num_classes])
logits = tf.reduce_mean(logits, [1, 2])
predictions = tf.argmax(logits, 1)
self.assertEquals(predictions.get_shape().as_list(), [eval_batch_size])
def testTrainEvalWithReuse(self):
train_batch_size = 2
eval_batch_size = 1
train_height, train_width = 224, 224
eval_height, eval_width = 300, 400
num_classes = 1000
with self.test_session():
train_inputs = tf.random.uniform(
(train_batch_size, train_height, train_width, 3))
logits, _ = alexnet.alexnet_v2(train_inputs)
self.assertListEqual(logits.get_shape().as_list(),
[train_batch_size, num_classes])
tf.compat.v1.get_variable_scope().reuse_variables()
eval_inputs = tf.random.uniform(
(eval_batch_size, eval_height, eval_width, 3))
logits, _ = alexnet.alexnet_v2(eval_inputs, is_training=False,
spatial_squeeze=False)
self.assertListEqual(logits.get_shape().as_list(),
[eval_batch_size, 4, 7, num_classes])
logits = tf.reduce_mean(input_tensor=logits, axis=[1, 2])
predictions = tf.argmax(input=logits, axis=1)
self.assertEquals(predictions.get_shape().as_list(), [eval_batch_size])
def testEndPoints(self):
batch_size = 5
height, width = 224, 224
num_classes = 1000
with self.test_session():
inputs = tf.random_uniform((batch_size, height, width, 3))
_, end_points = alexnet.alexnet_v2(inputs, num_classes)
expected_names = [
'alexnet_v2/conv1', 'alexnet_v2/pool1', 'alexnet_v2/conv2',
'alexnet_v2/pool2', 'alexnet_v2/conv3', 'alexnet_v2/conv4',
'alexnet_v2/conv5', 'alexnet_v2/pool5', 'alexnet_v2/fc6',
'alexnet_v2/fc7', 'alexnet_v2/fc8'
]
self.assertSetEqual(set(end_points.keys()), set(expected_names))
def build_train_op(image_tensor, label_tensor, is_training):
alexnet_argscope = alexnet_v2_arg_scope(weight_decay=FLAGS.weight_decay)
global_step = tf.get_variable(name="global_step", shape=[], dtype=tf.int32, trainable=False)
with slim.arg_scope(alexnet_argscope):
logits, end_points = alexnet_v2(image_tensor, is_training=is_training, num_classes=100)
loss = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits=logits, labels=label_tensor))
accuracy = tf.reduce_sum(tf.cast(tf.equal(tf.cast(tf.argmax(logits,1),tf.int32), label_tensor),tf.int32))
end_points['loss'], end_points['accuracy'] = loss, accuracy
if is_training:
optimizer = tf.train.AdadeltaOptimizer(learning_rate=FLAGS.learning_rate)
train_op = optimizer.minimize(loss, global_step=global_step)
return train_op, end_points
else:
return None, end_points
def main():
"""
You can also run these commands manually to generate the pb file
1. git clone https://github.com/tensorflow/models.git
2. export PYTHONPATH=Path_to_your_model_folder
3. python alexnet.py
"""
height, width = 224, 224
inputs = tf.Variable(tf.random_uniform((1, height, width, 3)), name = 'input')
net, end_points = alexnet.alexnet_v2(inputs, is_training=False)
print("nodes in the graph")
for n in end_points:
print(n + " => " + str(end_points[n]))
net_outputs = map(lambda x: tf.get_default_graph().get_tensor_by_name(x), argv[2].split(','))
run_model(net_outputs, argv[1], 'alexnet', argv[3] == 'True')
def load_model(model='Mobile'):
global x, y, sess
# 占位符
x = tf.placeholder(tf.float32, [None, IMG_SIZE, IMG_SIZE, 1])
# 模型保存路径，前向传播
if model == 'Alex':
log_path = 'weight/Alex'
y, _ = alexnet.alexnet_v2(
x,
num_classes=CLASSES, # 分类的类别
is_training=False, # 是否在训练
dropout_keep_prob=1.0, # 保留比率
spatial_squeeze=True, # 压缩掉1维的维度
global_pool=GLOBAL_POOL) # 输入不是规定的尺寸时，需要global_pool
elif model == 'Mobile':
log_path = 'weight/Mobile'
y, _ = mobilenet_v1.mobilenet_v1(x,
num_classes=CLASSES,
dropout_keep_prob=1.0,
is_training=False,
min_depth=8,
depth_multiplier=1.0,
conv_defs=None,
prediction_fn=None,
spatial_squeeze=True,
reuse=None,
scope='MobilenetV1',
global_pool=GLOBAL_POOL)
else:
print('Error: model name not exist')
return
y = tf.nn.softmax(y)
saver = tf.train.Saver()
sess = tf.Session()
# 恢复模型权重
# print('Reading checkpoints from: ', model)
ckpt = tf.train.get_checkpoint_state(log_path)
if ckpt and ckpt.model_checkpoint_path:
global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
saver.restore(sess, ckpt.model_checkpoint_path)
# print('Loading success, global_step is %s' % global_step)
else:
print('Error: no checkpoint file found')
return -1
def testNoClasses(self):
batch_size = 5
height, width = 224, 224
num_classes = None
with self.test_session():
inputs = tf.random_uniform((batch_size, height, width, 3))
net, end_points = alexnet.alexnet_v2(inputs, num_classes)
expected_names = [
'alexnet_v2/conv1', 'alexnet_v2/pool1', 'alexnet_v2/conv2',
'alexnet_v2/pool2', 'alexnet_v2/conv3', 'alexnet_v2/conv4',
'alexnet_v2/conv5', 'alexnet_v2/pool5', 'alexnet_v2/fc6',
'alexnet_v2/fc7'
]
self.assertSetEqual(set(end_points.keys()), set(expected_names))
self.assertTrue(net.op.name.startswith('alexnet_v2/fc7'))
self.assertListEqual(net.get_shape().as_list(),
[batch_size, 1, 1, 4096])
def main():
"""
You can also run these commands manually to generate the pb file
1. git clone https://github.com/tensorflow/models.git
2. export PYTHONPATH=Path_to_your_model_folder
3. python alexnet.py
"""
tf.set_random_seed(1)
height, width = 224, 224
inputs = tf.Variable(tf.random_uniform((1, height, width, 3)), name = 'input')
inputs = tf.identity(inputs, "input_node")
net, end_points = alexnet.alexnet_v2(inputs, is_training=False)
print("nodes in the graph")
for n in end_points:
print(n + " => " + str(end_points[n]))
net_outputs = map(lambda x: tf.get_default_graph().get_tensor_by_name(x), argv[2].split(','))
run_model(net_outputs, argv[1], 'alexnet', argv[3] == 'True')
def testEndPoints(self):
batch_size = 5
height, width = 224, 224
num_classes = 1000
with self.test_session():
inputs = tf.random_uniform((batch_size, height, width, 3))
_, end_points = alexnet.alexnet_v2(inputs, num_classes)
expected_names = ['alexnet_v2/conv1',
'alexnet_v2/pool1',
'alexnet_v2/conv2',
'alexnet_v2/pool2',
'alexnet_v2/conv3',
'alexnet_v2/conv4',
'alexnet_v2/conv5',
'alexnet_v2/pool5',
'alexnet_v2/fc6',
'alexnet_v2/fc7',
'alexnet_v2/fc8'
]
self.assertSetEqual(set(end_points.keys()), set(expected_names))
def load_model(model='Alex'):
global x, y, sess
# 占位符
x = tf.placeholder(tf.float32, [None, IMG_SIZE, IMG_SIZE, 3])
# 模型保存路径，前向传播
if model == 'Alex':
log_path = 'weight/Alex'
y, _ = alexnet.alexnet_v2(x,
num_classes=CLASSES, # 分类的类别
is_training=True, # 是否在训练
dropout_keep_prob=1.0, # 保留比率
spatial_squeeze=True, # 压缩掉1维的维度
global_pool=GLOBAL_POOL) # 输入不是规定的尺寸时，需要global_pool
elif model == 'VGG':
log_path = 'weight/VGG'
y, _ = vgg.vgg_16(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
spatial_squeeze=True,
global_pool=GLOBAL_POOL)
else:
print('Error: model name not exist')
return
y = tf.nn.softmax(y)
saver = tf.train.Saver()
sess = tf.Session()
# 恢复模型权重
print('Reading checkpoints from: ', model)
ckpt = tf.train.get_checkpoint_state(log_path)
if ckpt and ckpt.model_checkpoint_path:
global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
saver.restore(sess, ckpt.model_checkpoint_path)
print('Loading success, global_step is %s' % global_step)
else:
print('Error: no checkpoint file found')
return -1
def testNoClasses(self):
batch_size = 5
height, width = 224, 224
num_classes = None
with self.test_session():
inputs = tf.random_uniform((batch_size, height, width, 3))
net, end_points = alexnet.alexnet_v2(inputs, num_classes)
expected_names = ['alexnet_v2/conv1',
'alexnet_v2/pool1',
'alexnet_v2/conv2',
'alexnet_v2/pool2',
'alexnet_v2/conv3',
'alexnet_v2/conv4',
'alexnet_v2/conv5',
'alexnet_v2/pool5',
'alexnet_v2/fc6',
'alexnet_v2/fc7'
]
self.assertSetEqual(set(end_points.keys()), set(expected_names))
self.assertTrue(net.op.name.startswith('alexnet_v2/fc7'))
self.assertListEqual(net.get_shape().as_list(),
[batch_size, 1, 1, 4096])
def model_fn(features,labels,mode,params):
global_step = tf.train.get_global_step()
inputs = features['MRI']
logits, end_points= alexnet.alexnet_v2(inputs=inputs,
num_classes=30)
# predict mode
predicted_classes = tf.argmax(logits, 1)
if mode == tf.estimator.ModeKeys.PREDICT:
predictions = {
'class':predicted_classes,
'prob':tf.nn.softmax(logits)
}
return tf.estimator.EstimatorSpec(mode,predictions=predictions)
with tf.name_scope('accuracy'):
accuracy = tf.metrics.accuracy(labels=labels,predictions=predicted_classes)
my_acc = tf.reduce_mean(tf.cast(tf.equal(labels, predicted_classes), tf.float32))
tf.summary.scalar('accuracy',my_acc)
# compute loss
loss = tf.losses.sparse_softmax_cross_entropy(labels=labels,logits=logits)
# hook
train_hook_list = []
train_tensors_log = {'accuracy':accuracy[1],
'my_acc': my_acc,
'loss':loss,
'global_step':global_step}
train_hook_list.append(tf.train.LoggingTensorHook(tensors=train_tensors_log,every_n_iter=200))
# training op
if mode == tf.estimator.ModeKeys.TRAIN:
optimizer = tf.train.MomentumOptimizer(learning_rate=0.001,momentum=0.9)
train_op = optimizer.minimize(loss,global_step=tf.train.get_global_step())
return tf.estimator.EstimatorSpec(mode=mode,loss=loss,train_op=train_op,training_hooks=train_hook_list)
# compute evaluation metrics
eval_metric_ops = {
'accuracy':tf.metrics.accuracy(labels=labels,predictions=predicted_classes)
}
return tf.estimator.EstimatorSpec(mode=mode,loss=loss,eval_metric_ops=eval_metric_ops)
def predict(models_path, image_dir, labels_nums, data_format):
[batch_size, resize_height, resize_width, depths] = data_format
# labels = np.loadtxt(labels_filename, str, delimiter='\t')
input_images = tf.placeholder(
dtype=tf.float32,
shape=[None, resize_height, resize_width, depths],
name='input')
# Define the model:
with slim.arg_scope(alexnet.alexnet_v2_arg_scope()):
out, end_points = alexnet.alexnet_v2(inputs=input_images,
num_classes=labels_nums,
dropout_keep_prob=1.0,
is_training=False)
# 将输出结果进行softmax分布,再求最大概率所属类别
sess = tf.InteractiveSession()
sess.run(tf.global_variables_initializer())
saver = tf.train.Saver()
saver.restore(sess, models_path)
images_list = glob.glob(os.path.join(image_dir, '*.jpg'))
for image_path in images_list:
im = read_image(image_path,
resize_height,
resize_width,
normalization=True)
im = im[np.newaxis, :]
#pred = sess.run(f_cls, feed_dict={x:im, keep_prob:1.0})
pre_score = sess.run([out], feed_dict={input_images: im})
mean_std = statistic.get_score(pre_score, type="mean_std")
# print("image_path:{},pre_score:{},mean_std:{}".format(image_path,pre_score,mean_std))
print("image_path:{},mean_std:{}".format(image_path, mean_std))
sess.close()
def predict(img_path, model='Alex'):
"""
预测图片
:param img_path: 图片路径
:param model: 模型名
:return: none
"""
# 占位符
x = tf.placeholder(tf.float32, [None, IMG_SIZE, IMG_SIZE, 3])
# 模型保存路径，前向传播
if model == 'Alex':
log_path = "../log/Alex"
y, _ = alexnet.alexnet_v2(x,
num_classes=CLASSES, # 分类的类别
is_training=True, # 是否在训练
dropout_keep_prob=1.0, # 保留比率
spatial_squeeze=True, # 压缩掉1维的维度
global_pool=GLOBAL_POOL) # 输入不是规定的尺寸时，需要global_pool
elif model == 'VGG':
log_path = "../log/VGG"
y, _ = vgg.vgg_16(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
spatial_squeeze=True,
global_pool=GLOBAL_POOL)
elif model == 'Incep4':
log_path = "../log/Incep4"
y, _ = inception_v4.inception_v4(x, num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
reuse=None,
scope='InceptionV4',
create_aux_logits=True)
elif model == 'Res':
log_path = "../log/Res"
y, _ = resnet_v2.resnet_v2_50(x,
num_classes=CLASSES,
is_training=True,
global_pool=GLOBAL_POOL,
output_stride=None,
spatial_squeeze=True,
reuse=None,
scope='resnet_v2_50')
else:
print('Error: model name not exist')
return
y = tf.nn.softmax(y)
saver = tf.train.Saver()
with tf.Session() as sess:
# 恢复模型权重
print('Reading checkpoints: ', model)
ckpt = tf.train.get_checkpoint_state(log_path)
if ckpt and ckpt.model_checkpoint_path:
global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
saver.restore(sess, ckpt.model_checkpoint_path)
print('Loading success, global_step is %s' % global_step)
else:
print('Error: no checkpoint file found')
return
img = read_img(img_path)
# 第一次运行时间会较长
sess.run(y, feed_dict={x: img})
start_time = time.clock()
predictions = sess.run(y, feed_dict={x: img})
pre = np.argmax(predictions)
end_time = time.clock()
runtime = end_time - start_time
print('prediction is:', predictions)
print('predict class is:', pre)
print('run time:', runtime)
def classify_image(filepath):
with tf.Graph().as_default():
image = open(filepath, 'rb')
# Open specified url and load image as a string
image_string = image.read()
# Decode string into matrix with intensity values
image = tf.image.decode_jpeg(image_string, channels=3)
# Resize the input image, preserving the aspect ratio
# and make a central crop of the resulted image.
# The crop will be of the size of the default image size of
# the network.
processed_image = vgg_preprocessing.preprocess_image(image,
image_size,
image_size,
is_training=False)
# Networks accept images in batches.
# The first dimension usually represents the batch size.
# In our case the batch size is one.
processed_images = tf.expand_dims(processed_image, 0)
# Create the model, use the default arg scope to configure
# the batch norm parameters. arg_scope is a very convenient
# feature of slim library -- you can define default
# parameters for layers -- like stride, padding etc.
with slim.arg_scope(alexnet.alexnet_v2_arg_scope()):
logits, _ = alexnet.alexnet_v2(processed_images,
num_classes=6,
is_training=False)
# In order to get probabilities we apply softmax on the output.
probabilities = tf.nn.softmax(logits)
# Create a function that reads the network weights
# from the checkpoint file that you downloaded.
# We will run it in session later.
init_fn = slim.assign_from_checkpoint_fn(
os.path.join(checkpoints_dir, 'model.ckpt-100000'),
slim.get_model_variables('alexnet_v2'))
with tf.Session() as sess:
# Load weights
init_fn(sess)
# We want to get predictions, image as numpy matrix
# and resized and cropped piece that is actually
# being fed to the network.
np_image, network_input, probabilities = sess.run(
[image, processed_image, probabilities])
probabilities = probabilities[0, 0:]
sorted_inds = [
i[0]
for i in sorted(enumerate(-probabilities), key=lambda x: x[1])
]
for i in range(6):
index = sorted_inds[i]
print('Probability %0.2f => [%s]' %
(probabilities[index], names[index]))
return sorted_inds[0], probabilities
# tensorflow.python.framework.errors_impl.OutOfRangeError: RandomShuffleQueue '_0_shuffle_batch/random_shuffle_queue' is closed and has insufficient elements (requested 25, current size 0)
#定义网络结构
#调用修改过的alexnet_v2
# train_network_fn = nets_factory.get_network_fn(
# 'alexnet_v2',
# num_classes=CHAR_SET_LEN,
# weight_decay=0.0005,
# is_training=True)
# train_network_fn = alexnet.alexnet_v2(num_classes=CHAR_SET_LEN,is_training=True)
with tf.Session() as sess:
# inputs: a tensor of size [batch_size, height, width, channels]
X = tf.reshape(x, [BATCH_SIZE, 224, 224, 1])
# 数据输入网络得到输出值
logits0,logits1,logits2,logits3,end_points = alexnet.alexnet_v2(inputs=X,num_classes=CHAR_SET_LEN,is_training=True) #train_network_fn(X)
# 把标签转成one_hot的形式
one_hot_labels0 = tf.one_hot(indices=tf.cast(y0, tf.int32), depth=CHAR_SET_LEN)
one_hot_labels1 = tf.one_hot(indices=tf.cast(y1, tf.int32), depth=CHAR_SET_LEN)
one_hot_labels2 = tf.one_hot(indices=tf.cast(y2, tf.int32), depth=CHAR_SET_LEN)
one_hot_labels3 = tf.one_hot(indices=tf.cast(y3, tf.int32), depth=CHAR_SET_LEN)
# 计算loss
#用sigmoid_cross_entropy_with_logits代替softmax_cross_entropy_with_logits_v2，更合理一些，这与alexnet实现有关
loss0 = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(logits=logits0,labels=one_hot_labels0))
loss1 = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(logits=logits1,labels=one_hot_labels1))
loss2 = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(logits=logits2,labels=one_hot_labels2))
loss3 = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(logits=logits3,labels=one_hot_labels3))
# loss0 = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits0,labels=one_hot_labels0))
# loss1 = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits1,labels=one_hot_labels1))
# loss2 = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits2,labels=one_hot_labels2))
# loss3 = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits3,labels=one_hot_labels3))
def _construct_model(model_type='resnet_v1_50'):
"""Constructs model for the desired type of CNN.
Args:
model_type: Type of model to be used.
Returns:
end_points: A dictionary from components of the network to the corresponding
activations.
Raises:
ValueError: If the model_type is not supported.
"""
# Placeholder input.
images = array_ops.placeholder(
dtypes.float32, shape=(1, None, None, 3), name=_INPUT_NODE)
# Construct model.
if model_type == 'inception_resnet_v2':
_, end_points = inception.inception_resnet_v2_base(images)
elif model_type == 'inception_resnet_v2-same':
_, end_points = inception.inception_resnet_v2_base(
images, align_feature_maps=True)
elif model_type == 'inception_v2':
_, end_points = inception.inception_v2_base(images)
elif model_type == 'inception_v2-no-separable-conv':
_, end_points = inception.inception_v2_base(
images, use_separable_conv=False)
elif model_type == 'inception_v3':
_, end_points = inception.inception_v3_base(images)
elif model_type == 'inception_v4':
_, end_points = inception.inception_v4_base(images)
elif model_type == 'alexnet_v2':
_, end_points = alexnet.alexnet_v2(images)
elif model_type == 'vgg_a':
_, end_points = vgg.vgg_a(images)
elif model_type == 'vgg_16':
_, end_points = vgg.vgg_16(images)
elif model_type == 'mobilenet_v1':
_, end_points = mobilenet_v1.mobilenet_v1_base(images)
elif model_type == 'mobilenet_v1_075':
_, end_points = mobilenet_v1.mobilenet_v1_base(
images, depth_multiplier=0.75)
elif model_type == 'resnet_v1_50':
_, end_points = resnet_v1.resnet_v1_50(
images, num_classes=None, is_training=False, global_pool=False)
elif model_type == 'resnet_v1_101':
_, end_points = resnet_v1.resnet_v1_101(
images, num_classes=None, is_training=False, global_pool=False)
elif model_type == 'resnet_v1_152':
_, end_points = resnet_v1.resnet_v1_152(
images, num_classes=None, is_training=False, global_pool=False)
elif model_type == 'resnet_v1_200':
_, end_points = resnet_v1.resnet_v1_200(
images, num_classes=None, is_training=False, global_pool=False)
elif model_type == 'resnet_v2_50':
_, end_points = resnet_v2.resnet_v2_50(
images, num_classes=None, is_training=False, global_pool=False)
elif model_type == 'resnet_v2_101':
_, end_points = resnet_v2.resnet_v2_101(
images, num_classes=None, is_training=False, global_pool=False)
elif model_type == 'resnet_v2_152':
_, end_points = resnet_v2.resnet_v2_152(
images, num_classes=None, is_training=False, global_pool=False)
elif model_type == 'resnet_v2_200':
_, end_points = resnet_v2.resnet_v2_200(
images, num_classes=None, is_training=False, global_pool=False)
else:
raise ValueError('Unsupported model_type %s.' % model_type)
return end_points
graph = tf.Graph()
with graph.as_default():
tf.logging.set_verbosity(tf.logging.INFO)
dataset, images, labels = getImageBatchAndOneHotLabels(
dataset_dir, 'train', num_readers, num_preprocessing_threads,
batch_size)
_, images_val, labels_val = getImageBatchAndOneHotLabels(
dataset_dir, 'validation', 2, 2, batch_size)
# Create Model network and endpoints
with slim.arg_scope(alexnet.alexnet_v2_arg_scope()):
logits, _ = alexnet.alexnet_v2(images, num_classes=dataset.num_classes)
with tf.variable_scope(tf.get_variable_scope(), reuse=True):
logits_val, _ = alexnet.alexnet_v2(images_val,
num_classes=dataset.num_classes)
#Metrics
accuracy_validation = slim.metrics.accuracy(
tf.to_int32(tf.argmax(logits_val, 1)),
tf.to_int32(tf.argmax(labels_val, 1)))
top5_accuracy = tf.metrics.mean(
tf.nn.in_top_k(logits_val, tf.to_int32(tf.argmax(labels_val, 1)), k=5))
# Added Loss Function
tf.losses.softmax_cross_entropy(labels, logits)
total_loss = slim.losses.get_total_loss()
checkpoints_dir = './network1/model.ckpt-4832'
path_unnormal = './CROP_IMG/unnormal/'
path_normal = './CROP_IMG/normal/'
image_names = os.listdir(path_unnormal)
image_names.sort()
tf.reset_default_graph()
images = tf.placeholder(tf.float32, [1, 224, 224, 3], name='input_images')
with slim.arg_scope(alexnet.alexnet_v2_arg_scope()):
logits, end_points1 = alexnet.alexnet_v2(images,
num_classes=2,
is_training=False)
probabilities1 = tf.nn.softmax(logits)
init_fn = slim.assign_from_checkpoint_fn(checkpoints_dir,
slim.get_variables_to_restore(),
ignore_missing_vars=True)
sess = tf.Session()
init_fn(sess)
for j, image_name in enumerate(image_names):
print('unnormal:', j)
image_path = os.path.join(path_unnormal, image_name)
def train(inherit=False, model='Alex'):
# 加载数据集
images = np.load('../data/phone_data.npy')
labels = np.load('../data/phone_label.npy')
train_data, val_data, train_label, val_label = train_test_split(
images, labels, test_size=0.2, random_state=222)
# 占位符
x = tf.placeholder(tf.float32, [None, IMG_SIZE, IMG_SIZE, 1],
name='x-input')
y_ = tf.placeholder(tf.float32, [None, CLASSES], name='y-input')
my_global_step = tf.Variable(0,
name='global_step',
trainable=False,
dtype=tf.int64)
# 前向传播
if model == 'Alex':
log_path = "../log/Alex"
model_name = 'alex.ckpt'
y, _ = alexnet.alexnet_v2(
x,
num_classes=CLASSES, # 分类的类别
is_training=True, # 是否在训练
dropout_keep_prob=1.0, # 保留比率
spatial_squeeze=True, # 压缩掉1维的维度
global_pool=GLOBAL_POOL) # 输入不是规定的尺寸时，需要global_pool
else:
log_path = '../log/My'
model_name = 'my.ckpt'
y, _ = lenet5(x)
# 交叉熵、损失值、优化器、准确率
loss = tf.reduce_mean(tf.sqrt(tf.square(y_ - y) + 0.0000001))
learning_rate = tf.train.exponential_decay(LEARNING_RATE_BASE,
my_global_step, 100,
LEARNING_RATE_DECAY)
train_op = tf.train.GradientDescentOptimizer(learning_rate).minimize(
loss, global_step=my_global_step)
# 模型保存器、初始化
saver = tf.train.Saver()
init = tf.global_variables_initializer()
tf.summary.scalar("loss", loss)
# tf.summary.histogram("W1", W)
merged_summary_op = tf.summary.merge_all()
# 训练迭代
with tf.Session() as sess:
summary_writer1 = tf.summary.FileWriter('../log/curve/train',
sess.graph)
summary_writer2 = tf.summary.FileWriter('../log/curve/test')
step = 0
if inherit:
ckpt = tf.train.get_checkpoint_state(log_path)
if ckpt and ckpt.model_checkpoint_path:
global_step = ckpt.model_checkpoint_path.split('/')[-1].split(
'-')[-1]
saver.restore(sess, ckpt.model_checkpoint_path)
print(model, 'continue train from %s:' % global_step)
step = int(global_step)
else:
print('Error: no checkpoint file found')
return
else:
print(model, 'restart train:')
sess.run(init)
# 迭代
while step < MAX_STEP:
start_time = time.clock()
image_batch, label_batch = utils.get_batch(train_data, train_label,
BATCH_SIZE)
# 训练，损失值和准确率
_, train_loss = sess.run([train_op, loss],
feed_dict={
x: image_batch,
y_: label_batch
})
end_time = time.clock()
runtime = end_time - start_time
step += 1
# 训练信息、曲线和保存模型
if step % INFO_STEP == 0 or step == MAX_STEP:
summary_str = sess.run(merged_summary_op,
feed_dict={
x: image_batch,
y_: label_batch
})
summary_writer1.add_summary(summary_str, step)
test_loss, summary_str = sess.run([loss, merged_summary_op],
feed_dict={
x: val_data,
y_: val_label
})
summary_writer2.add_summary(summary_str, step)
print(
'step: %d, runtime: %.2f, train loss: %.4f, test loss: %.4f'
% (step, runtime, train_loss, test_loss))
if step % SAVE_STEP == 0:
checkpoint_path = os.path.join(log_path, model_name)
saver.save(sess, checkpoint_path, global_step=step)
def evaluate(model='Alex'):
"""
评估模型
:param model: model name
:return: none
"""
# 预测为某类的样本个数，某类预测正确的样本个数
pre_pos = np.zeros(CLASSES, dtype=np.uint16)
true_pos = np.zeros(CLASSES, dtype=np.uint16)
# 加载测试集
images = np.load('../data/data_299.npy')
labels = np.load('../data/label_299.npy')
_, val_data, _, val_label = train_test_split(images,
labels,
test_size=0.2,
random_state=222)
test_num = val_data.shape[0]
# 存放预测结果
res = np.zeros(test_num, dtype=np.uint16)
# 占位符
x = tf.placeholder(tf.float32, [None, IMG_SIZE, IMG_SIZE, 3],
name="x_input")
y_ = tf.placeholder(tf.uint8, [None], name="y_input")
# 模型保存路径，模型名，预训练文件路径，前向传播
if model == 'Alex':
log_dir = "../log/Alex"
y, _ = alexnet.alexnet_v2(
x,
num_classes=CLASSES, # 分类的类别
is_training=True, # 是否在训练
dropout_keep_prob=1.0, # 保留比率
spatial_squeeze=True, # 压缩掉1维的维度
global_pool=GLOBAL_POOL) # 输入不是规定的尺寸时，需要global_pool
elif model == 'VGG':
log_dir = "../log/VGG"
y, _ = vgg.vgg_16(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
spatial_squeeze=True,
global_pool=GLOBAL_POOL)
elif model == 'Incep4':
log_dir = "E:/alum/log/Incep4"
y, _ = inception_v4.inception_v4(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
reuse=None,
scope='InceptionV4',
create_aux_logits=True)
elif model == 'Res':
log_dir = "E:/alum/log/Res"
y, _ = resnet_v2.resnet_v2_50(x,
num_classes=CLASSES,
is_training=True,
global_pool=GLOBAL_POOL,
output_stride=None,
spatial_squeeze=True,
reuse=None,
scope='resnet_v2_50')
else:
print('Error: model name not exist')
return
# 预测结果
y_pre = tf.argmax(y, 1)
saver = tf.train.Saver(tf.global_variables())
with tf.Session() as sess:
# 恢复模型权重
print('Reading checkpoints of', model)
ckpt = tf.train.get_checkpoint_state(log_dir)
if ckpt and ckpt.model_checkpoint_path:
global_step = ckpt.model_checkpoint_path.split('/')[-1].split(
'-')[-1]
saver.restore(sess, ckpt.model_checkpoint_path)
print('Loading success, step is %s' % global_step)
else:
print('Error: no checkpoint file found')
return
# 遍历一次测试集需要次数
num_iter = int(math.ceil(test_num / test_num))
step = 0
start = 0
while step < num_iter:
# 获取一个 batch
if step == num_iter - 1:
end = test_num
else:
end = start + BATCH_SIZE
image_batch = val_data[start:end]
label_batch = val_label[start:end]
# 准确率和预测结果统计信息
pre = sess.run(y_pre, feed_dict={x: image_batch, y_: label_batch})
pre = np.squeeze(pre)
res[start:end] = pre
start += BATCH_SIZE
step += 1
# 计算准确率
normal_num = 0
for i in range(test_num):
pre_pos[res[i]] += 1
if res[i] == val_label[i]:
true_pos[res[i]] += 1
if val_label[i] == 0:
normal_num += 1
precision = true_pos / pre_pos
print('测试样本数：', test_num)
print('预测数：', pre_pos)
print('正确数：', true_pos)
print('各类精度：', precision)
print('准确率：', np.sum(true_pos) / test_num)
print('平均准确率：', np.mean(precision))
print('漏检率：', (pre_pos[0] - true_pos[0]) / (test_num - normal_num))
print('过杀率：', (normal_num - true_pos[0]) / normal_num)
def submit(test_dir, model='VGG'):
"""
测试集预测结果保存为 csv 文件
:param test_dir: 测试集根路径
:param model: 模型名称
:return: none
"""
# 类别名称
print('running submit')
tf.reset_default_graph()
class_label = [
'norm', 'defect1', 'defect2', 'defect3', 'defect4', 'defect5',
'defect6', 'defect7', 'defect8', 'defect9', 'defect10', 'defect11'
]
img_index, res = [], []
# 占位符
x = tf.placeholder(tf.float32, [None, IMG_SIZE, IMG_SIZE, 3])
# 模型保存路径，前向传播
if model == 'Alex':
log_dir = "../log/Alex"
y, _ = alexnet.alexnet_v2(
x,
num_classes=CLASSES, # 分类的类别
is_training=True, # 是否在训练
dropout_keep_prob=1.0, # 保留比率
spatial_squeeze=True, # 压缩掉1维的维度
global_pool=GLOBAL_POOL) # 输入不是规定的尺寸时，需要global_pool
elif model == 'VGG':
log_dir = "../log/VGG"
y, _ = vgg.vgg_16(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
spatial_squeeze=True,
global_pool=GLOBAL_POOL)
elif model == 'Incep4':
log_dir = "E:/alum/log/Incep4"
y, _ = inception_v4.inception_v4(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
reuse=None,
scope='InceptionV4',
create_aux_logits=True)
elif model == 'Res':
log_dir = "E:/alum/log/Res"
y, _ = resnet_v2.resnet_v2_50(x,
num_classes=CLASSES,
is_training=True,
global_pool=GLOBAL_POOL,
output_stride=None,
spatial_squeeze=True,
reuse=None,
scope='resnet_v2_50')
else:
print('Error: model name not exist')
return
saver = tf.train.Saver()
with tf.Session() as sess:
# 恢复模型权重
print("Reading checkpoints...")
# ckpt 有 model_checkpoint_path 和 all_model_checkpoint_paths 两个属性
ckpt = tf.train.get_checkpoint_state(log_dir)
if ckpt and ckpt.model_checkpoint_path:
global_step = ckpt.model_checkpoint_path.split('/')[-1].split(
'-')[-1]
saver.restore(sess, ckpt.model_checkpoint_path)
print('Loading success, global_step is %s' % global_step)
else:
print('Error: no checkpoint file found')
return
img_list = os.listdir(test_dir)
for img_name in img_list:
# print(img_name)
# 读取图片并缩放
img = Image.open(os.path.join(test_dir, img_name))
img = img.resize((IMG_SIZE, IMG_SIZE))
# 图片转为numpy的数组
img = np.array(img, np.float32)
img = np.expand_dims(img, axis=0)
predictions = sess.run(y, feed_dict={x: img})
pre = np.argmax(predictions)
img_index.append(img_name)
res.append(class_label[int(pre)])
print('image: ', img_index, '\n', 'class:', res)
# 存为csv文件。两列：图片名，预测结果
data = pd.DataFrame({'0': img_index, '1': res})
# data.sort_index(axis=0)
data.to_csv("../submit/vgg_2.csv", index=False, header=False)
def _construct_model(model_type='resnet_v1_50'):
"""Constructs model for the desired type of CNN.
Args:
model_type: Type of model to be used.
Returns:
end_points: A dictionary from components of the network to the corresponding
activations.
Raises:
ValueError: If the model_type is not supported.
"""
# Placeholder input.
images = array_ops.placeholder(dtypes.float32,
shape=(1, None, None, 3),
name=_INPUT_NODE)
# Construct model.
if model_type == 'inception_resnet_v2':
_, end_points = inception.inception_resnet_v2_base(images)
elif model_type == 'inception_resnet_v2-same':
_, end_points = inception.inception_resnet_v2_base(
images, align_feature_maps=True)
elif model_type == 'inception_v2':
_, end_points = inception.inception_v2_base(images)
elif model_type == 'inception_v2-no-separable-conv':
_, end_points = inception.inception_v2_base(images,
use_separable_conv=False)
elif model_type == 'inception_v3':
_, end_points = inception.inception_v3_base(images)
elif model_type == 'inception_v4':
_, end_points = inception.inception_v4_base(images)
elif model_type == 'alexnet_v2':
_, end_points = alexnet.alexnet_v2(images)
elif model_type == 'vgg_a':
_, end_points = vgg.vgg_a(images)
elif model_type == 'vgg_16':
_, end_points = vgg.vgg_16(images)
elif model_type == 'mobilenet_v1':
_, end_points = mobilenet_v1.mobilenet_v1_base(images)
elif model_type == 'mobilenet_v1_075':
_, end_points = mobilenet_v1.mobilenet_v1_base(images,
depth_multiplier=0.75)
elif model_type == 'resnet_v1_50':
_, end_points = resnet_v1.resnet_v1_50(images,
num_classes=None,
is_training=False,
global_pool=False)
elif model_type == 'resnet_v1_101':
_, end_points = resnet_v1.resnet_v1_101(images,
num_classes=None,
is_training=False,
global_pool=False)
elif model_type == 'resnet_v1_152':
_, end_points = resnet_v1.resnet_v1_152(images,
num_classes=None,
is_training=False,
global_pool=False)
elif model_type == 'resnet_v1_200':
_, end_points = resnet_v1.resnet_v1_200(images,
num_classes=None,
is_training=False,
global_pool=False)
elif model_type == 'resnet_v2_50':
_, end_points = resnet_v2.resnet_v2_50(images,
num_classes=None,
is_training=False,
global_pool=False)
elif model_type == 'resnet_v2_101':
_, end_points = resnet_v2.resnet_v2_101(images,
num_classes=None,
is_training=False,
global_pool=False)
elif model_type == 'resnet_v2_152':
_, end_points = resnet_v2.resnet_v2_152(images,
num_classes=None,
is_training=False,
global_pool=False)
elif model_type == 'resnet_v2_200':
_, end_points = resnet_v2.resnet_v2_200(images,
num_classes=None,
is_training=False,
global_pool=False)
else:
raise ValueError('Unsupported model_type %s.' % model_type)
return end_points
def train(train_filename, train_images_dir, train_log_step, train_param,
val_filename, val_images_dir, val_log_step, labels_nums, data_shape,
snapshot, snapshot_prefix):
'''
:param train_record_file: 训练的tfrecord文件
:param train_log_step: 显示训练过程log信息间隔
:param train_param: train参数
:param val_record_file: 验证的tfrecord文件
:param val_log_step: 显示验证过程log信息间隔
:param val_param: val参数
:param labels_nums: labels数
:param data_shape: 输入数据shape
:param snapshot: 保存模型间隔
:param snapshot_prefix: 保存模型文件的前缀名
:return:
'''
[base_lr, max_steps] = train_param
[batch_size, resize_height, resize_width, depths] = data_shape
# # 从record中读取图片和labels数据
tf_image, tf_labels = read_images(train_filename,
train_images_dir,
data_shape,
shuffle=True,
type='normalization')
train_images_batch, train_labels_batch = get_batch_images(
tf_image,
tf_labels,
batch_size=batch_size,
labels_nums=labels_nums,
one_hot=False,
shuffle=True)
# Define the model:
with slim.arg_scope(alexnet.alexnet_v2_arg_scope()):
out, end_points = alexnet.alexnet_v2(inputs=input_images,
num_classes=labels_nums,
dropout_keep_prob=keep_prob,
is_training=is_training)
loss = tf.reduce_sum(tf.squared_difference(x=out, y=input_labels))
# loss1=tf.squared_difference(x=out,y=input_labels)
# loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=output, labels=y))
train_op = tf.train.AdamOptimizer(learning_rate=base_lr).minimize(loss)
# tf.losses.add_loss(loss1)
# # slim.losses.add_loss(my_loss)
# loss = tf.losses.get_total_loss(add_regularization_losses=True) # 添加正则化损失loss=2.2
# # Specify the optimization scheme:
# optimizer = tf.train.GradientDescentOptimizer(learning_rate=base_lr)
# # create_train_op that ensures that when we evaluate it to get the loss,
# # the update_ops are done and the gradient updates are computed.
# train_op = slim.learning.create_train_op(total_loss=loss, optimizer=optimizer)
saver = tf.train.Saver()
max_acc = 0.0
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
sess.run(tf.local_variables_initializer())
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(sess=sess, coord=coord)
for i in range(max_steps + 1):
batch_input_images, batch_input_labels = sess.run(
[train_images_batch, train_labels_batch])
_, train_loss = sess.run(
[train_op, loss],
feed_dict={
input_images: batch_input_images,
input_labels: batch_input_labels,
keep_prob: 0.5,
is_training: True
})
if i % train_log_step == 0:
print("%s: Step [%d] train Loss : %f" %
(datetime.now(), i, train_loss))
# # train测试(这里仅测试训练集的一个batch)
# if i%train_log_step == 0:
# train_acc = sess.run(accuracy, feed_dict={input_images:batch_input_images,
# input_labels: batch_input_labels,
# keep_prob:1.0, is_training: False})
# print "%s: Step [%d] train Loss : %f, training accuracy : %g" % (datetime.now(), i, train_loss, train_acc)
#
# # val测试(测试全部val数据)
# if i%val_log_step == 0:
# _, train_loss = sess.run([train_step, loss], feed_dict={input_images: batch_input_images,
# input_labels: batch_input_labels,
# keep_prob: 1.0, is_training: False})
# print "%s: Step [%d] val Loss : %f, val accuracy : %g" % (datetime.now(), i, mean_loss, mean_acc)
#
# 模型保存:每迭代snapshot次或者最后一次保存模型
if (i % snapshot == 0 and i > 0) or i == max_steps:
print('-----save:{}-{}'.format(snapshot_prefix, i))
saver.save(sess, snapshot_prefix, global_step=i)
# # 保存val准确率最高的模型
# if mean_acc>max_acc and mean_acc>0.5:
# max_acc=mean_acc
# path = os.path.dirname(snapshot_prefix)
# best_models=os.path.join(path,'best_models_{}_{:.4f}.ckpt'.format(i,max_acc))
# print('------save:{}'.format(best_models))
# saver.save(sess, best_models)
coord.request_stop()
coord.join(threads)
def train(model='Alex', inherit=False, fine_tune=False):
"""
train a specified model
:param model: the train model name
:param inherit: whether to continue training
:param fine_tune: whether to fine tuning
:return: none
"""
# 占位符
x = tf.placeholder(tf.float32, [None, IMG_SIZE, IMG_SIZE, 3],
name="x_input")
y_ = tf.placeholder(tf.uint8, [None], name="y_input")
my_global_step = tf.Variable(0,
name='global_step',
trainable=False,
dtype=tf.int64)
# 加载数据集
images = np.load('../data/data_' + str(IMG_SIZE) + '.npy')
labels = np.load('../data/label_' + str(IMG_SIZE) + '.npy')
train_data, val_data, train_label, val_label = train_test_split(
images, labels, test_size=0.2, random_state=222)
# 模型保存路径，模型名，预训练文件路径，前向传播
if model == 'Alex':
log_path = "../log/Alex"
model_name = 'alex.ckpt'
if fine_tune:
print('Error: alex has no pre-train model')
return
y, _ = alexnet.alexnet_v2(
x,
num_classes=CLASSES, # 分类的类别
is_training=True, # 是否在训练
dropout_keep_prob=1.0, # 保留比率
spatial_squeeze=True, # 压缩掉1维的维度
global_pool=GLOBAL_POOL) # 输入不是规定的尺寸时，需要global_pool
elif model == 'VGG':
log_path = "../log/VGG"
model_name = 'vgg_299.ckpt'
fine_tune_path = '../data/vgg_16_2016_08_28/vgg_16.ckpt'
y, _ = vgg.vgg_16(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=0.8,
spatial_squeeze=True,
global_pool=GLOBAL_POOL)
variables_to_restore = slim.get_variables_to_restore(
exclude=['vgg_16/fc8'])
elif model == 'Incep':
log_path = "E:/alum/log/Incep"
model_name = 'incep.ckpt'
fine_tune_path = 'E:/alum/weight/inception_v4_2016_09_09/inception_v4.ckpt'
y, _ = inception_v4.inception_v4(x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=0.8,
reuse=None,
scope='InceptionV4',
create_aux_logits=True)
variables_to_restore = slim.get_variables_to_restore(
exclude=['InceptionV4/Logits', 'InceptionV4/AuxLogits'])
elif model == 'Res':
log_path = "E:/alum/log/Res"
model_name = 'res.ckpt'
fine_tune_path = 'E:/alum/weight/resnet_v2_50_2017_04_14/resnet_v2_50.ckpt'
y, _ = resnet_v2.resnet_v2_50(x,
num_classes=CLASSES,
is_training=True,
global_pool=GLOBAL_POOL,
output_stride=None,
spatial_squeeze=True,
reuse=None,
scope='resnet_v2_50')
variables_to_restore = slim.get_variables_to_restore(
exclude=['resnet_v2_50/logits'])
elif model == 'IncepRes':
log_path = "E:/alum/log/IncepRes"
model_name = 'incepres.ckpt'
fine_tune_path = 'E:/alum/weight/inception_resnet_v2_2016_08_30/inception_resnet_v2_2016_08_30.ckpt'
y, _ = inception_resnet_v2.inception_resnet_v2(
x,
num_classes=CLASSES,
is_training=True,
dropout_keep_prob=1.0,
reuse=None,
scope='InceptionResnetV2',
create_aux_logits=True,
activation_fn=tf.nn.relu)
variables_to_restore = slim.get_variables_to_restore(exclude=[
'InceptionResnetV2/Logits', 'InceptionResnetV2/AuxLogits'
])
else:
print('Error: model name not exist')
return
y_hot = tf.one_hot(y_, CLASSES)
cross_entropy = tf.nn.softmax_cross_entropy_with_logits(logits=y,
labels=y_hot,
name='entropy')
loss = tf.reduce_mean(cross_entropy, name='loss')
learning_rate = tf.train.exponential_decay(LEARNING_RATE_BASE,
my_global_step, 100,
LEARNING_RATE_DECAY)
optimizer = tf.train.AdamOptimizer(learning_rate)
train_op = optimizer.minimize(loss, global_step=my_global_step)
correct = tf.equal(tf.argmax(y, 1), tf.argmax(y_hot, 1))
accuracy = tf.reduce_mean(tf.cast(correct, "float"))
# 模型保存的Saver
saver = tf.train.Saver(tf.global_variables())
init = tf.global_variables_initializer()
tf.summary.scalar("loss", loss)
# tf.summary.histogram("W1", W)
merged_summary_op = tf.summary.merge_all()
# 训练迭代
with tf.Session() as sess:
step = 0
train_summary_writer = tf.summary.FileWriter('../log/curve/train',
sess.graph)
test_summary_writer = tf.summary.FileWriter('../log/curve/test')
# 是否使用预训练模型
if fine_tune:
print('finetune:', model)
# 预训练模型的Saver
init_fine_tune = slim.assign_from_checkpoint_fn(
fine_tune_path, variables_to_restore, ignore_missing_vars=True)
sess.run(init)
init_fine_tune(sess)
# 是否继续训练，ckpt 有 model_checkpoint_path 和 all_model_checkpoint_paths 两个属性
elif inherit:
ckpt = tf.train.get_checkpoint_state(log_path)
if ckpt and ckpt.model_checkpoint_path:
global_step = ckpt.model_checkpoint_path.split('/')[-1].split(
'-')[-1]
saver.restore(sess, ckpt.model_checkpoint_path)
print(model, 'continue train from %s:' % global_step)
step = int(global_step)
else:
print('Error: no checkpoint file found')
return
else:
print(model, 'restart train:')
sess.run(init)
# 迭代
while step < MAX_STEP:
start_time = time.clock()
image_batch, label_batch = utils.get_batch(train_data, train_label,
BATCH_SIZE)
# 如果输入是灰色图，要增加一维
# image_batch = np.expand_dims(image_batch, axis=3)
# 训练，损失值和准确率
_ = sess.run(train_op, feed_dict={x: image_batch, y_: label_batch})
end_time = time.clock()
runtime = end_time - start_time
step += 1
# 训练信息和保存模型
if step % INFO_STEP == 0 or step == MAX_STEP:
train_loss, train_ac, train_summary = sess.run(
[loss, accuracy, merged_summary_op],
feed_dict={
x: image_batch,
y_: label_batch
})
train_summary_writer.add_summary(train_summary, step)
test_loss, test_ac, test_summary = sess.run(
[loss, accuracy, merged_summary_op],
feed_dict={
x: val_data,
y_: val_label
})
test_summary_writer.add_summary(test_summary, step)
print(
'step: %d, runtime: %.2f, train_loss: %.4f, test_loss: %.4f,train accuracy: %.4f, '
'test accuracy: %.4f' %
(step, runtime, train_loss, test_loss, train_ac, test_ac))
if step % SAVE_STEP == 0:
print('learning_rate: ', sess.run(learning_rate))
checkpoint_path = os.path.join(log_path, model_name)
saver.save(sess, checkpoint_path, global_step=step)
def train(run_dir,
master,
task_id,
num_readers,
from_graspnet_checkpoint,
dataset_dir,
checkpoint_dir,
save_summaries_steps,
save_interval_secs,
num_preprocessing_threads,
num_steps,
hparams,
scope='graspnet'):
for path in [run_dir]:
if not tf.gfile.Exists(path):
tf.gfile.Makedirs(path)
hparams_filename = os.path.join(run_dir, 'hparams.json')
with tf.gfile.FastGFile(hparams_filename, 'w') as f:
f.write(hparams.to_json())
with tf.Graph().as_default():
with tf.device(tf.train.replica_device_setter(task_id)):
global_step = slim.get_or_create_global_step()
images, class_labels, theta_labels = get_dataset(
dataset_dir, num_readers, num_preprocessing_threads, hparams)
'''
with slim.arg_scope(vgg.vgg_arg_scope()):
net, end_points = vgg.vgg_16(inputs=images,
num_classes=num_classes,
is_training=True,
dropout_keep_prob=0.7,
scope=scope)
'''
with slim.arg_scope(alexnet.alexnet_v2_arg_scope()):
net, end_points = alexnet.alexnet_v2(inputs=images,
num_classes=num_classes,
is_training=True,
dropout_keep_prob=0.7,
scope=scope)
loss, accuracy = create_loss(net, class_labels, theta_labels)
learning_rate = hparams.learning_rate
if hparams.lr_decay_step:
learning_rate = tf.train.exponential_decay(
hparams.learning_rate,
slim.get_or_create_global_step(),
decay_steps=hparams.lr_decay_step,
decay_rate=hparams.lr_decay_rate,
staircase=True)
tf.summary.scalar('Learning_rate', learning_rate)
optimizer = tf.train.GradientDescentOptimizer(learning_rate)
train_op = slim.learning.create_train_op(loss, optimizer)
add_summary(images, end_points, loss, accuracy, scope=scope)
summary_op = tf.summary.merge_all()
variable_map = restore_map(
from_graspnet_checkpoint=from_graspnet_checkpoint,
scope=scope,
model_name=hparams.model_name,
checkpoint_exclude_scope='fc8')
init_saver = tf.train.Saver(variable_map)
def initializer_fn(sess):
init_saver.restore(sess, checkpoint_dir)
tf.logging.info('Successfully load pretrained checkpoint.')
init_fn = initializer_fn
session_config = tf.ConfigProto(allow_soft_placement=True,
log_device_placement=False)
session_config.gpu_options.allow_growth = True
saver = tf.train.Saver(
keep_checkpoint_every_n_hours=save_interval_secs,
max_to_keep=100)
slim.learning.train(
train_op,
logdir=run_dir,
master=master,
global_step=global_step,
session_config=session_config,
# init_fn=init_fn,
summary_op=summary_op,
number_of_steps=num_steps,
startup_delay_steps=15,
save_summaries_secs=save_summaries_steps,
saver=saver)
