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))
vgg.vgg_16(inputs, num_classes)
expected_names = [
'vgg_16/conv1/conv1_1/weights',
'vgg_16/conv1/conv1_1/biases',
'vgg_16/conv1/conv1_2/weights',
'vgg_16/conv1/conv1_2/biases',
'vgg_16/conv2/conv2_1/weights',
'vgg_16/conv2/conv2_1/biases',
'vgg_16/conv2/conv2_2/weights',
'vgg_16/conv2/conv2_2/biases',
'vgg_16/conv3/conv3_1/weights',
'vgg_16/conv3/conv3_1/biases',
'vgg_16/conv3/conv3_2/weights',
'vgg_16/conv3/conv3_2/biases',
'vgg_16/conv3/conv3_3/weights',
'vgg_16/conv3/conv3_3/biases',
'vgg_16/conv4/conv4_1/weights',
'vgg_16/conv4/conv4_1/biases',
'vgg_16/conv4/conv4_2/weights',
'vgg_16/conv4/conv4_2/biases',
'vgg_16/conv4/conv4_3/weights',
'vgg_16/conv4/conv4_3/biases',
'vgg_16/conv5/conv5_1/weights',
'vgg_16/conv5/conv5_1/biases',
'vgg_16/conv5/conv5_2/weights',
'vgg_16/conv5/conv5_2/biases',
'vgg_16/conv5/conv5_3/weights',
'vgg_16/conv5/conv5_3/biases',
'vgg_16/fc6/weights',
'vgg_16/fc6/biases',
'vgg_16/fc7/weights',
'vgg_16/fc7/biases',
'vgg_16/fc8/weights',
'vgg_16/fc8/biases',
]
model_variables = [v.op.name for v in slim.get_model_variables()]
self.assertSetEqual(set(model_variables), set(expected_names))
def testFullyConvolutional(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, _ = vgg.vgg_16(inputs, num_classes, spatial_squeeze=False)
self.assertEquals(logits.op.name, 'vgg_16/fc8/BiasAdd')
self.assertListEqual(logits.get_shape().as_list(),
[batch_size, 2, 2, num_classes])
def total_loss(inputs, stylized_inputs, style_gram_matrices, content_weights,
style_weights, reuse=False):
"""Computes the total loss function.
The total loss function is composed of a content, a style and a total
variation term.
Args:
inputs: Tensor. The input images.
stylized_inputs: Tensor. The stylized input images.
style_gram_matrices: dict mapping layer names to their corresponding
Gram matrices.
content_weights: dict mapping layer names to their associated content loss
weight. Keys that are missing from the dict won't have their content
loss computed.
style_weights: dict mapping layer names to their associated style loss
weight. Keys that are missing from the dict won't have their style
loss computed.
reuse: bool. Whether to reuse model parameters. Defaults to False.
Returns:
Tensor for the total loss, dict mapping loss names to losses.
"""
# Propagate the input and its stylized version through VGG16.
end_points = vgg.vgg_16(inputs, reuse=reuse)
stylized_end_points = vgg.vgg_16(stylized_inputs, reuse=True)
# Compute the content loss
total_content_loss, content_loss_dict = content_loss(
end_points, stylized_end_points, content_weights)
# Compute the style loss
total_style_loss, style_loss_dict = style_loss(
style_gram_matrices, stylized_end_points, style_weights)
# Compute the total loss
loss = total_content_loss + total_style_loss
loss_dict = {'total_loss': loss}
loss_dict.update(content_loss_dict)
loss_dict.update(style_loss_dict)
return loss, loss_dict
def tower_loss(data_tensor, label_tensor, num_classes, train_mode):
#vgg = tf.contrib.slim.nets.vgg
with slim.arg_scope(vgg.vgg_arg_scope(weight_decay=args.weight_decay)):
logits, endpoints_dict = vgg.vgg_16(data_tensor, num_classes=num_classes, is_training=train_mode,dropout_keep_prob=args.dropout_keep_prob)
loss=tf.losses.sparse_softmax_cross_entropy(labels=label_tensor, logits=logits)
return loss, logits
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, _ = vgg.vgg_16(inputs, num_classes)
self.assertEquals(logits.op.name, 'vgg_16/fc8/squeezed')
self.assertListEqual(logits.get_shape().as_list(),
[batch_size, 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, _ = vgg.vgg_16(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 CRAFT_net(inputs, is_trianing=True, reuse=None, weight_decay=0.9):
with slim.arg_scope(vgg_arg_scope()):
vgg_res, end_points = vgg_16(inputs)
with tf.variable_scope('vgg_16', [end_points.values]):
batch_norm_params = {
'decay': 0.997,
'epsilon': 1e-5,
'scale': True,
'is_training': is_trianing
}
with slim.arg_scope(
[slim.conv2d],
activation_fn=tf.nn.relu,
normalizer_fn=slim.batch_norm,
normalizer_params=batch_norm_params,
weights_regularizer=slim.l2_regularizer(weight_decay)):
f = [
end_points['vgg_16/conv2/conv2_2'],
end_points['vgg_16/conv3/conv3_3'],
end_points['vgg_16/conv4/conv4_3'],
end_points['vgg_16/conv5/conv5_3']
]
net = f[3]
# VGG end
net = slim.max_pool2d(net, [3, 3],
stride=1,
padding='SAME',
scope='pool5') # w/16 512
net = arous_conv(net, 3, 3, 1024, 6,
name='arous_conv') # w/16 1024
net = slim.conv2d(net, 1024, [1, 1], padding='SAME',
scope='conv6') # w/16 1024
# U-net start
net = tf.concat([net, f[3]], axis=3) # w/16 1024 + 512
net = upconvBlock(net, 512, 256) # w/16 256
net = upsample(net, (64, 64))
net = tf.concat([net, f[2]], axis=3) # w/8 256 + 512
net = upconvBlock(net, 256, 128) # w/8 128
net = upsample(net, (128, 128))
net = tf.concat([net, f[1]], axis=3) # w/4 128 + 256
net = upconvBlock(net, 128, 64) # w/4 64
net = upsample(net, (256, 256))
net = tf.concat([net, f[0]], axis=3) # w/2 64 + 128
net = upconvBlock(net, 64, 32) # w/2 32
# U-net end
net = slim.repeat(net, 2, slim.conv2d, 32, [3, 3]) # w/2 32
net = slim.conv2d(net, 16, [3, 3], padding='SAME') # w/2 16
net = slim.conv2d(net, 16, [1, 1], padding='SAME') # w/2 16
net = slim.conv2d(net, 2, [1, 1], padding='SAME') # w/2 2
return net, end_points
def build_model(self):
# get content_img, style_img and define gen_img
if content_input is not None:
self.content_path = content_input
if style_input is not None:
self.style_path = style_input
content_img, content_shape = utils.load_content_img(self.content_path)
style_img = utils.load_style_img(self.style_path, content_shape)
content_img_shape = content_img.shape
gen_img = tf.Variable(tf.random_normal(content_img_shape) * 0.256)
with slim.arg_scope(vgg.vgg_arg_scope()):
f1, f2, f3, f4, exclude = vgg.vgg_16(
tf.concat([gen_img, content_img, style_img], axis=0))
# calculate content_loss and style_loss
content_loss = utils.cal_content_loss(f3)
style_loss = utils.cal_style_loss(f1, f2, f3, f4)
# load vgg model
vgg_model_path = VGG_MODEL_PATH
vgg_vars = slim.get_variables_to_restore(include=['vgg_16'],
exclude=exclude)
init_fn = slim.assign_from_checkpoint_fn(vgg_model_path, vgg_vars)
init_fn(self.sess)
print('vgg_16 weights load done')
self.gen_img = gen_img
self.global_step = tf.Variable(0,
name='global_step',
trainable=False)
self.content_loss = content_loss
self.style_loss = style_loss * W_STYLE
# the total loss
self.loss = self.content_loss + self.style_loss
# starter_learning_rate = 1e1
# global_step = tf.train.get_global_step()
# learning_rate = tf.train.exponential_decay(starter_learning_rate, global_step, decay_steps=500,
# decay_rate=0.98,
# staircase=True)
self.opt = tf.train.AdamOptimizer(LEARNING_RATE).minimize(
self.loss, global_step=self.global_step, var_list=gen_img)
all_var = tf.global_variables()
init_var = [v for v in all_var if 'vgg_16' not in v.name]
init = tf.variables_initializer(var_list=init_var)
self.sess.run(init)
self.save = tf.train.Saver()
def vgg_16(inputs, no_fc=False):
with slim.arg_scope(vgg.vgg_arg_scope()):
net, end_points = vgg.vgg_16(inputs,
None,
is_training=False,
spatial_squeeze=False,
fc_conv_padding='SAME',
no_fc=no_fc)
if no_fc:
return end_points['vgg_16/pool5'], end_points[
'vgg_16/pool4'], end_points['vgg_16/pool3']
else:
return net, end_points['vgg_16/pool4'], end_points['vgg_16/pool3']
def _fc7_flat(eval_dataset):
inputs = eval_dataset.get_next("inputs")
logits, end_points = vgg.vgg_16(
inputs,
num_classes=1000,
is_training=False, # this time, we don't train the network.
dropout_keep_prob=0.5, # doesn't matter since "is_training=False."
spatial_squeeze=True, # squeeze spatial dimensions in the final output
scope='vgg_16',
fc_conv_padding=
'VALID', # this code uses conv instead of fc (they are equivalent!)
global_pool=False)
fc7_features = end_points['vgg_16/fc7']
return tf.layers.flatten(fc7_features, name="fc7_flat")
def vgg_encoding(self, processed_images, is_training, reuse=False):
with slim.arg_scope(vgg.vgg_arg_scope()):
fc7 = vgg.vgg_16(processed_images,
num_classes=self.no_classes,
is_training=is_training,
spatial_squeeze=False,
fc_conv_padding='VALID',
reuse=reuse,
return_fc7=True,
fc7_size=self.fc7_size)
return fc7
def arch_vgg16(X, num_classes, dropout_keep_prob=0.8, is_train=False):
arg_scope = vgg_arg_scope()
with slim.arg_scope(arg_scope):
net, end_points = vgg_16(X, is_training=is_train)
with slim.arg_scope([slim.conv2d, slim.max_pool2d, slim.avg_pool2d],
stride=1,
padding='SAME'):
with tf.variable_scope('Logits_out'):
net = slim.conv2d(net,
num_classes, [1, 1],
activation_fn=None,
normalizer_fn=None,
scope='fc8')
net = tf.squeeze(net, [1, 2], name='fc8/squeezed')
return net
def testTrainEvalWithReuse(self):
train_batch_size = 2
eval_batch_size = 1
train_height, train_width = 224, 224
eval_height, eval_width = 256, 256
num_classes = 1000
with self.test_session():
train_inputs = tf.random_uniform(
(train_batch_size, train_height, train_width, 3))
logits, _ = vgg.vgg_16(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, _ = vgg.vgg_16(eval_inputs,
is_training=False,
spatial_squeeze=False)
self.assertListEqual(logits.get_shape().as_list(),
[eval_batch_size, 2, 2, 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 net_feats(ims, num_classes=None, is_training=True, reuse=False, scope=''):
if net_type == 'resnet':
feats = resnet_v1.resnet_v1_50(ims,
num_classes=num_classes,
reuse=reuse,
is_training=is_training,
scope='%s_resnet_v1_50' % scope)[0]
feats = slim.flatten(feats)
elif net_type == 'vgg':
feats = vgg.vgg_16(ims,
num_classes=None,
scope='%s_vgg_16' % scope,
reuse=reuse,
is_training=is_training)[0]
print 'CNN feature shape:', shape(feats)
return feats
def build_model(self):
# get content_img, style_img and define gen_img
if content_input is not None:
self.content_path = content_input
if style_input is not None:
self.style_path = style_input
content_img, content_shape = utils.load_content_img(self.content_path)
style_img = utils.load_style_img(self.style_path, content_shape)
content_img_shape = content_img.shape
gen_img = tf.Variable(tf.random_normal(content_img_shape) * 0.256)
with slim.arg_scope(vgg.vgg_arg_scope()):
f1, f2, f3, f4, exclude = vgg.vgg_16(tf.concat([gen_img, content_img, style_img], axis=0))
# calculate content_loss and style_loss
content_loss = utils.cal_content_loss(f3)
style_loss = utils.cal_style_loss(f1, f2, f3, f4)
# load vgg model
vgg_model_path = VGG_MODEL_PATH
vgg_vars = slim.get_variables_to_restore(include=['vgg_16'], exclude=exclude)
init_fn = slim.assign_from_checkpoint_fn(vgg_model_path, vgg_vars)
init_fn(self.sess)
print('vgg_16 weights load done')
self.gen_img = gen_img
self.global_step = tf.Variable(0, name='global_step', trainable=False)
self.content_loss = content_loss
self.style_loss = style_loss * W_STYLE
# the total loss
self.loss = self.content_loss + self.style_loss
# starter_learning_rate = 1e1
# global_step = tf.train.get_global_step()
# learning_rate = tf.train.exponential_decay(starter_learning_rate, global_step, decay_steps=500,
# decay_rate=0.98,
# staircase=True)
self.opt = tf.train.AdamOptimizer(LEARNING_RATE).minimize(self.loss, global_step=self.global_step,
var_list=gen_img)
all_var = tf.global_variables()
init_var = [v for v in all_var if 'vgg_16' not in v.name]
init = tf.variables_initializer(var_list=init_var)
self.sess.run(init)
self.save = tf.train.Saver()
def precompute_gram_matrices(image, final_endpoint='fc8'):
"""Pre-computes the Gram matrices on a given image.
Args:
image: 4-D tensor. Input (batch of) image(s).
final_endpoint: str, name of the final layer to compute Gram matrices for.
Defaults to 'fc8'.
Returns:
dict mapping layer names to their corresponding Gram matrices.
"""
with tf.Session() as session:
end_points = vgg.vgg_16(image, final_endpoint=final_endpoint)
tf.train.Saver(slim.get_variables('vgg_16')).restore(
session, vgg.checkpoint_file())
return dict((key, gram_matrix(value).eval())
for key, value in end_points.items())
def _buildGraph(self):
x_in = tf.placeholder(
tf.float32,
shape=[
None, # enables variable batch size
self.input_dim[0]
],
name="x")
x_in_reshape = tf.reshape(
x_in, [-1, self.input_dim[1], self.input_dim[2], 3])
dropout = tf.placeholder_with_default(1., shape=[], name="dropout")
y_in = tf.placeholder(dtype=tf.int8, name="y")
onehot_labels = tf.one_hot(indices=tf.cast(y_in, tf.int32), depth=2)
is_train = tf.placeholder_with_default(True, shape=[], name="is_train")
logits, nett, ww = vgg.vgg_16(x_in_reshape,
num_classes=2,
is_training=is_train,
dropout_keep_prob=dropout,
spatial_squeeze=True,
scope='vgg16')
pred = tf.nn.softmax(logits, name="prediction")
global_step = tf.Variable(0, trainable=False)
pred_cost = tf.losses.softmax_cross_entropy(
onehot_labels=onehot_labels, logits=logits)
tf.summary.scalar("InceptionV3_cost", pred_cost)
train_op = tf.contrib.layers.optimize_loss(
loss=pred_cost,
learning_rate=self.learning_rate,
global_step=global_step,
optimizer="Adam")
merged_summary = tf.summary.merge_all()
return (x_in, dropout, is_train, y_in, logits, nett, ww, pred,
pred_cost, global_step, train_op, merged_summary)
def test_with_origin():
image=cv2.imread('./imgdata/cat.18.jpg')
print(image.shape)
res=cv2.resize(image,(224,224))
res_image=np.expand_dims(res,0)
print(res_image.shape,type(res_image))
graph = tf.Graph
inputs = tf.placeholder(dtype=tf.float32, shape=[None, 224, 224, 3], name='inputs')
net, end_points = vgg.vgg_16(inputs, num_classes=1000)
print(end_points)
saver=tf.train.Saver()
with tf.Session() as sess:
saver.restore(sess,'./models/vgg_16.ckpt')
input=sess.graph.get_tensor_by_name('inputs:0')
output=sess.graph.get_tensor_by_name('vgg_16/fc8/squeezed:0')
pred=sess.run(output,feed_dict={input:res_image})
print(np.argmax(pred,1))
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 = vgg.vgg_16(inputs, num_classes)
expected_names = [
'vgg_16/conv1/conv1_1', 'vgg_16/conv1/conv1_2', 'vgg_16/pool1',
'vgg_16/conv2/conv2_1', 'vgg_16/conv2/conv2_2', 'vgg_16/pool2',
'vgg_16/conv3/conv3_1', 'vgg_16/conv3/conv3_2',
'vgg_16/conv3/conv3_3', 'vgg_16/pool3', 'vgg_16/conv4/conv4_1',
'vgg_16/conv4/conv4_2', 'vgg_16/conv4/conv4_3', 'vgg_16/pool4',
'vgg_16/conv5/conv5_1', 'vgg_16/conv5/conv5_2',
'vgg_16/conv5/conv5_3', 'vgg_16/pool5', 'vgg_16/fc6',
'vgg_16/fc7', 'vgg_16/fc8'
]
self.assertSetEqual(set(end_points.keys()), set(expected_names))
def vgg16_fcn_net(image_tensor,
number_of_classes,
is_training=True,
upsample_factor=8):
# tf.reset_default_graph()
# Define the model that we want to use -- specify to use only two classes at the last layer
with slim.arg_scope(vgg.vgg_arg_scope()):
logits, end_points = vgg.vgg_16(image_tensor,
num_classes=number_of_classes,
is_training=is_training,
spatial_squeeze=False,
fc_conv_padding='SAME')
downsampled_logits_shape = tf.shape(logits)
img_shape = tf.shape(image_tensor)
# Calculate the ouput size of the upsampled tensor
# The shape should be batch_size X width X height X num_classes
upsampled_logits_shape = tf.stack([
downsampled_logits_shape[0], img_shape[1], img_shape[2],
downsampled_logits_shape[3]
])
# Perform the upsampling x2
upsampled_logits = upsample(logits, 'vgg_16/fc8/t_conv_x2', 2,
end_points['vgg_16/pool4'], 'conv_pool4',
number_of_classes)
# Perform the upsampling x2
upsampled_logits = upsample(upsampled_logits, 'vgg_16/fc8/t_conv_x2_x2', 2,
end_points['vgg_16/pool3'], 'conv_pool3',
number_of_classes)
# Perform the upsampling x8
upsample_filter_tensor_x8 = bilinear_upsample_weights(
upsample_factor, number_of_classes, 'vgg_16/fc8/t_conv_x8')
upsampled_logits = tf.nn.conv2d_transpose(
upsampled_logits,
upsample_filter_tensor_x8,
output_shape=upsampled_logits_shape,
strides=[1, upsample_factor, upsample_factor, 1],
padding='SAME')
return upsampled_logits
def __init__(self, learning_rate, num_classes=1000, is_training=True):
self.global_step = tf.Variable(0, trainable=False, name='global_step')
self.batch_imgs = tf.placeholder(tf.float32, (None, None, None, 3))
self.batch_lbls = tf.placeholder(tf.int32, (None, 1))
_, end_points = vgg_16(self.batch_imgs, num_classes, is_training)
self.loss = tf.losses.sparse_softmax_cross_entropy(
labels=self.batch_lbls, logits=end_points['vgg_16/fc8'])
gradients = tf.gradients(self.loss, tf.trainable_variables())
self.update = tf.train.GradientDescentOptimizer(
learning_rate).apply_gradients(zip(gradients,
tf.trainable_variables()),
global_step=self.global_step)
self.saver = tf.train.Saver(tf.global_variables(),
max_to_keep=99999999)
def vgg16_fcn8_model(images,
num_classes,
is_training=False,
raw_image_shape=(520 - 170, 800),
decoder='fcn8'):
train_image_shape = (224 * 2, 224 * 3)
if decoder == 'fcn8':
decoder_fn = mobilenet_v1_fcn_decoder
elif decoder == 'fcn8_upsample':
decoder_fn = mobilenet_v1_fcn8_upsample_decoder
else:
raise ValueError("the decoder should be either fcn8 or fcn8_upsample")
if images.dtype != tf.uint8:
raise ValueError("the image should be uint8")
images = tf.image.resize_images(images, size=train_image_shape)
tf.summary.image('input_image_after_rescale_and_resize',
tf.expand_dims(images[0], 0))
processed_images = tf.map_fn(vgg_preprocessing.vgg_image_rescale,
images,
dtype=tf.float32)
# Create the model, use the default arg scope to configure the batch norm parameters.
with slim.arg_scope(vgg.vgg_arg_scope()):
# 1000 classes instead of 1001.
logits, end_points = vgg.vgg_16(processed_images,
num_classes=1000,
is_training=is_training,
spatial_squeeze=False)
layer4 = end_points['vgg_16/pool3']
layer6 = end_points['vgg_16/pool4']
layer13 = end_points['vgg_16/pool5']
last_layer = decoder_fn(layer13, layer4, layer6, num_classes)
last_layer = post_process_logits(end_points, last_layer, raw_image_shape,
train_image_shape)
return last_layer, end_points
def finetune():
image=cv2.imread('./imgdata/cat.18.jpg')
print(image.shape)
res_image=cv2.resize(image,(224,224),interpolation=cv2.INTER_CUBIC)
print(res_image.shape)
res_image=np.expand_dims(res_image,axis=0)
print(res_image.shape)
labels=[[1,0]]
graph=tf.get_default_graph()
input=tf.placeholder(dtype=tf.float32,shape=[None,224,224,3],name='inputs')
y_=tf.placeholder(dtype=tf.float32,shape=[None,2],name='labels')
net,end_points=vgg.vgg_16(input,num_classes=2)
print(net,end_points)
y=tf.nn.softmax(net)
cross_entropy=tf.reduce_mean(-tf.reduce_sum(y_*tf.log(y),reduction_indices=[1]))
output_vars=tf.get_collection(tf.GraphKeys.TRAINABLE_VARIABLES,scope='vgg_16/fc8')
print(output_vars)
train_op=tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy,var_list=output_vars)
var=tf.global_variables()
print(var)
var_to_restore=[val for val in var if 'fc8' not in val.name]
print(var_to_restore)
saver=tf.train.Saver(var_to_restore)
with tf.Session() as sess:
saver.restore(sess,'./models/vgg_16.ckpt')
var_to_init=[val for val in var if 'fc8' in val.name]
sess.run(tf.variables_initializer(var_to_init))
w1=sess.graph.get_tensor_by_name('vgg_16/conv1/conv1_1/weights:0')
print(sess.run(w1,feed_dict={input:res_image}))
w8=sess.graph.get_tensor_by_name('vgg_16/fc8/weights:0')
print('w8',sess.run(w8,feed_dict={input:res_image}))
sess.run(train_op,feed_dict={input:res_image,y_:labels})
def build_network(self):
"""Builds the image model subgraph and generates image feature maps.
"""
# dimension convention: depth x height x width x channels
inputs = self.images
# with tf.variable_scope('model', [inputs]) as sc:
# if self.mode == "validation":
# sc.reuse_variables()
net, end_points = vgg_16(inputs, num_classes=self.num_classes, is_training=self.is_training())
# net = imagenet_resnet_v2(50, self.num_classes)(inputs, self.is_training())
print("self.num_classes")
print(self.num_classes)
print('complete network build.')
return net
def look_model_params():
graph=tf.Graph
inputs=tf.placeholder(dtype=tf.float32,shape=[None,224,224,3],name='inputs')
net,end_points=vgg.vgg_16(inputs,num_classes=1000)
saver=tf.train.Saver()
with tf.Session() as sess:
saver.restore(sess,'models/vgg_16.ckpt')
tvs=[v for v in tf.trainable_variables()]
print('获得所有可训练变量的权重:')
for v in tvs:
print(v.name)
print(sess.run(v))
gv=[v for v in tf.global_variables()]
print('获得所有变量:')
for v in gv:
print(v.name,'\n')
ops=[o for o in sess.graph.get_operations()]
print('获得所有operations相关的tensor:')
for o in ops:
print(o.name,'\n')
def build_model(self):
train_imgs = tools.load_train_img(TRAIN_DATA_DIR, self.batch_size, self.img_size)
style_imgs = tools.load_style_img(STYLE_IMAGE_PATH)
with slim.arg_scope(model.arg_scope()):
gen_img, variables = model.inference(train_imgs, reuse=False, name='transform')
with slim.arg_scope(vgg.vgg_arg_scope()):
gen_img_processed = [tf.image.per_image_standardization(image) for image in
tf.unstack(gen_img, axis=0, num=self.batch_size)]
f1, f2, f3, f4, exclude = vgg.vgg_16(tf.concat([gen_img_processed, train_imgs, style_imgs], axis=0))
gen_f, img_f, _ = tf.split(f4, 3, 0)
content_loss = tf.nn.l2_loss(gen_f - img_f) / tf.to_float(tf.size(gen_f))
style_loss = model.styleloss(f1, f2, f3, f4)
vgg_model_path = VGG_MODEL_PATH
vgg_vars = slim.get_variables_to_restore(include=['vgg_16'], exclude=exclude)
init_fn = slim.assign_from_checkpoint_fn(vgg_model_path, vgg_vars)
init_fn(self.sess)
print("vgg's weights load done")
self.gen_img = gen_img
self.global_step = tf.Variable(0, name="global_step", trainable=False)
self.content_loss = content_loss
self.style_loss = style_loss * self.style_w
self.loss = self.content_loss + self.style_loss
self.learn_rate = tf.train.exponential_decay(self.learn_rate_base, self.global_step, 1,
self.learn_rate_decay, staircase=True)
self.opt = tf.train.AdamOptimizer(self.learn_rate).minimize(self.loss, global_step=self.global_step,
var_list=variables)
all_var = tf.global_variables()
init_var = [v for v in all_var if 'vgg_16' not in v.name]
init = tf.variables_initializer(var_list=init_var)
self.sess.run(init)
self.save = tf.train.Saver(var_list=variables)
log_folder = os.path.join(FLAGS.output_dir, 'train')
vgg_checkpoint_path = FLAGS.checkpoint_path
# Creates a variable to hold the global_step.
global_step = tf.Variable(0,
trainable=False,
name='global_step',
dtype=tf.int64)
# Define the model that we want to use -- specify to use only two classes at the last layer
with slim.arg_scope(vgg.vgg_arg_scope()):
logits, end_points = vgg.vgg_16(image_tensor,
num_classes=number_of_classes,
is_training=is_training_placeholder,
spatial_squeeze=False,
fc_conv_padding='SAME')
downsampled_logits_shape = tf.shape(logits)
img_shape = tf.shape(image_tensor)
# Calculate the ouput size of the upsampled tensor
# The shape should be batch_size X width X height X num_classes
upsampled_logits_shape = tf.stack([
downsampled_logits_shape[0], img_shape[1], img_shape[2],
downsampled_logits_shape[3]
])
#取出pool4的结果
def init_model(self):
slim = tf.contrib.slim
# 初始化
tf.reset_default_graph()
# 输入图片
image_tensor = tf.placeholder(tf.float32,
shape=(1, None, None, 3),
name='image_tensor')
orig_img_tensor = tf.placeholder(tf.uint8,
shape=(1, None, None, 3),
name='orig_img_tensor')
# 初始化模型
with slim.arg_scope(vgg.vgg_arg_scope()):
logits, end_points = vgg.vgg_16(image_tensor,
num_classes=self.number_of_classes,
is_training=False,
spatial_squeeze=False,
fc_conv_padding='SAME')
downsampled_logits_shape = tf.shape(logits)
img_shape = tf.shape(image_tensor)
# Calculate the ouput size of the upsampled tensor
# The shape should be batch_size X width X height X num_classes
upsampled_logits_shape = tf.stack([
downsampled_logits_shape[0], img_shape[1], img_shape[2],
downsampled_logits_shape[3]
])
pool4_feature = end_points['vgg_16/pool4']
with tf.variable_scope('vgg_16/fc8'):
aux_logits_16s = slim.conv2d(
pool4_feature,
self.number_of_classes, [1, 1],
activation_fn=None,
weights_initializer=tf.zeros_initializer,
scope='conv_pool4')
# Perform the upsampling
upsample_filter_np_x2 = bilinear_upsample_weights(
2, # upsample_factor,
self.number_of_classes)
upsample_filter_tensor_x2 = tf.Variable(upsample_filter_np_x2,
name='vgg_16/fc8/t_conv_x2')
upsampled_logits = tf.nn.conv2d_transpose(
logits,
upsample_filter_tensor_x2,
output_shape=tf.shape(aux_logits_16s),
strides=[1, 2, 2, 1],
padding='SAME')
upsampled_logits = upsampled_logits + aux_logits_16s
upsample_filter_np_x16 = bilinear_upsample_weights(
self.upsample_factor, self.number_of_classes)
upsample_filter_tensor_x16 = tf.Variable(upsample_filter_np_x16,
name='vgg_16/fc8/t_conv_x16')
upsampled_logits = tf.nn.conv2d_transpose(
upsampled_logits,
upsample_filter_tensor_x16,
output_shape=upsampled_logits_shape,
strides=[1, self.upsample_factor, self.upsample_factor, 1],
padding='SAME')
# Tensor to get the final prediction for each pixel -- pay
# attention that we don't need softmax in this case because
# we only need the final decision. If we also need the respective
# probabilities we will have to apply softmax.
pred = tf.argmax(upsampled_logits, axis=3, name='predictions')
probabilities = tf.nn.softmax(upsampled_logits, name='probabilities')
# 恢复模型，从训练好的模型中恢复参数
checkpoint_path = tf.train.latest_checkpoint(self.ckpt)
assert checkpoint_path, "no checkpoint exist, cant perform predict."
variables_to_restore = slim.get_model_variables()
sess_config = tf.ConfigProto()
sess_config.gpu_options.allow_growth = True
sess = tf.Session(config=sess_config)
init_op = tf.global_variables_initializer()
init_local_op = tf.local_variables_initializer()
saver = tf.train.Saver(max_to_keep=1)
# Run the initializers.
sess.run(init_op)
sess.run(init_local_op)
saver.restore(sess, checkpoint_path)
logging.debug('checkpoint restored from [{0}]'.format(checkpoint_path))
return sess, pred, orig_img_tensor, probabilities
log_folder = os.path.join(FLAGS.output_dir, 'train')
vgg_checkpoint_path = FLAGS.checkpoint_path
# Creates a variable to hold the global_step.
global_step = tf.Variable(0,
trainable=False,
name='global_step',
dtype=tf.int64)
# Define the model that we want to use -- specify to use only two classes at the last layer
with slim.arg_scope(vgg.vgg_arg_scope()):
logits, end_points = vgg.vgg_16(
image_tensor,
num_classes=number_of_classes,
is_training=is_training_placeholder,
spatial_squeeze=False, #包含空间信息 所以不做squeezn
fc_conv_padding='SAME'
) #wangluo suojian wei 1/32 224*224 --- 7*7*num_classes
downsampled_logits_shape = tf.shape(logits)
img_shape = tf.shape(image_tensor)
# Calculate the ouput size of the upsampled tensor
# The shape should be batch_size X width X height X num_classes
###dui logits jinxing shangcaiyang
upsampled_logits_shape = tf.stack([
downsampled_logits_shape[0], #
img_shape[1],
img_shape[2],
raise ValueError('Only have dataset: hand, standard, casia')
if not args.one_shot:
x_train = np.concatenate((x_train, x_test), axis=1)
t_train = np.concatenate((t_train, t_test), axis=1)
x_train = x_train.reshape([-1, n_xl, n_xl, n_channels]).astype(np.float32)
x_test = x_test.reshape([-1, n_xl, n_xl, n_channels]).astype(np.float32)
t_train = t_train.reshape([-1, n_y])
t_test = t_test.reshape([-1, n_y])
#is_training = tf.placeholder(tf.bool, shape=[])
x = tf.placeholder(tf.float32, shape=[None, n_xl, n_xl, n_channels])
y = tf.placeholder(tf.float32, shape=[None, n_y])
optimizer = tf.train.AdamOptimizer(lr, beta1=0.5)
net = vgg.vgg_16(x, n_y)
y_ = net.outputs
net.print_paras()
net.print_layers()
loss = tf.nn.softmax_cross_entropy_with_logits(labels=y, logits=y_)
#infer = optimizer.minimize(loss)
grads = optimizer.compute_gradients(loss)
infer = optimizer.apply_gradients(grads)
saver = tf.train.Saver(max_to_keep=10)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
ckpt_file = tf.train.latest_checkpoint(result_path)
begin_epoch = 1
if ckpt_file is not None:
#!/usr/bin/python
# -*- coding: utf-8 -*-
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import tensorflow as tf
import vgg
inputs = tf.placeholder(tf.float32, (None, 224, 224, 3), name='inputs')
r, g, b = tf.split(axis=3, num_or_size_splits=3, value=inputs * 255.0)
VGG_MEAN = [103.939, 116.779, 123.68]
bgr = tf.concat(values=[b - VGG_MEAN[0], g - VGG_MEAN[1], r - VGG_MEAN[2]],
axis=3)
with tf.contrib.slim.arg_scope(vgg.vgg_arg_scope()):
fc8, endpoints = vgg.vgg_16(bgr, is_training=False)
for k, v in endpoints.iteritems():
print(k, v)
upsample_factor = 16
number_of_classes = 21
log_folder = os.path.join(FLAGS.output_dir, 'train')
vgg_checkpoint_path = FLAGS.checkpoint_path
# Creates a variable to hold the global_step.
global_step = tf.Variable(0, trainable=False, name='global_step', dtype=tf.int64)
# Define the model that we want to use -- specify to use only two classes at the last layer
with slim.arg_scope(vgg.vgg_arg_scope()):
logits, end_points = vgg.vgg_16(image_tensor,
num_classes=number_of_classes,
is_training=is_training_placeholder,
spatial_squeeze=False,
fc_conv_padding='SAME')
downsampled_logits_shape = tf.shape(logits)
img_shape = tf.shape(image_tensor)
# Calculate the ouput size of the upsampled tensor
# The shape should be batch_size X width X height X num_classes
upsampled_logits_shape = tf.stack([
downsampled_logits_shape[0],
img_shape[1],
img_shape[2],
downsampled_logits_shape[3]
])
