def mobilenet_conv_layers(input, batch_size, num_unrolls):
input = tf.to_float(input) - IMAGENET_MEAN
conv1_1 = conv2d('conv_1',
input,
num_filters=int(round(32 * 1)),
kernel_size=(3, 3),
padding='SAME',
stride=(2, 2),
activation=tf.nn.relu6,
batchnorm_enabled=False,
is_training=True,
l2_strength=0.0,
bias=0.0)
#self.__add_to_nodes([conv1_1])
############################################################################################
conv2_1_dw, conv2_1_pw = depthwise_separable_conv2d(
'conv_ds_2',
conv1_1,
width_multiplier=1,
num_filters=64,
kernel_size=(3, 3),
padding='SAME',
stride=(1, 1),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv2_1_dw, conv2_1_pw])
conv2_2_dw, conv2_2_pw = depthwise_separable_conv2d(
'conv_ds_3',
conv2_1_pw,
width_multiplier=1,
num_filters=128,
kernel_size=(3, 3),
padding='SAME',
stride=(2, 2),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv2_2_dw, conv2_2_pw])
############################################################################################
#with tf.variable_scope('conv1_skip'):
if 0:
pool2 = tf.nn.max_pool(conv2_2_pw,
ksize=[1, 3, 3, 1],
strides=[1, 2, 2, 1],
padding='VALID',
name='pool2')
lrn2 = tf.nn.local_response_normalization(pool2,
depth_radius=2,
alpha=2e-5,
beta=0.75,
bias=1.0,
name='norm2')
prelu_skip = tf_util.get_variable('prelu',
shape=[16],
dtype=tf.float32,
initializer=prelu_initializer)
conv1_skip = tf_util.prelu(
tf_util.conv_layer(lrn2, 16, 1, activation=None), prelu_skip)
conv1_skip = tf.transpose(conv1_skip, perm=[0, 3, 1, 2])
conv1_skip_flat = tf_util.remove_axis(conv1_skip, [2, 3])
conv3_1_dw, conv3_1_pw = depthwise_separable_conv2d(
'conv_ds_4',
conv2_2_pw,
width_multiplier=1,
num_filters=128,
kernel_size=(3, 3),
padding='SAME',
stride=(1, 1),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv3_1_dw, conv3_1_pw])
conv3_2_dw, conv3_2_pw = depthwise_separable_conv2d(
'conv_ds_5',
conv3_1_pw,
width_multiplier=1,
num_filters=256,
kernel_size=(3, 3),
padding='SAME',
stride=(2, 2),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv3_2_dw, conv3_2_pw])
############################################################################################
conv4_1_dw, conv4_1_pw = depthwise_separable_conv2d(
'conv_ds_6',
conv3_2_pw,
width_multiplier=1,
num_filters=256,
kernel_size=(3, 3),
padding='SAME',
stride=(1, 1),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv4_1_dw, conv4_1_pw])
conv4_2_dw, conv4_2_pw = depthwise_separable_conv2d(
'conv_ds_7',
conv4_1_pw,
width_multiplier=1,
num_filters=512,
kernel_size=(3, 3),
padding='SAME',
stride=(2, 2),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv4_2_dw, conv4_2_pw])
############################################################################################
if 0:
pool3 = tf.nn.max_pool(conv4_2_pw,
ksize=[1, 3, 3, 1],
strides=[1, 2, 2, 1],
padding='VALID',
name='pool2')
lrn3 = tf.nn.local_response_normalization(pool2,
depth_radius=2,
alpha=2e-5,
beta=0.75,
bias=1.0,
name='norm2')
with tf.variable_scope('conv2_skip'):
prelu_skip = tf_util.get_variable('prelu',
shape=[16],
dtype=tf.float32,
initializer=prelu_initializer)
conv2_skip = tf_util.prelu(
tf_util.conv_layer(lrn3, 16, 1, activation=None), prelu_skip)
conv2_skip = tf.transpose(conv2_skip, perm=[0, 3, 1, 2])
conv2_skip_flat = tf_util.remove_axis(conv2_skip, [2, 3])
conv5_1_dw, conv5_1_pw = depthwise_separable_conv2d(
'conv_ds_8',
conv4_2_pw,
width_multiplier=1,
num_filters=512,
kernel_size=(3, 3),
padding='SAME',
stride=(1, 1),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv5_1_dw, conv5_1_pw])
conv5_2_dw, conv5_2_pw = depthwise_separable_conv2d(
'conv_ds_9',
conv5_1_pw,
width_multiplier=1,
num_filters=512,
kernel_size=(3, 3),
padding='SAME',
stride=(1, 1),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv5_2_dw, conv5_2_pw])
conv5_3_dw, conv5_3_pw = depthwise_separable_conv2d(
'conv_ds_10',
conv5_2_pw,
width_multiplier=1,
num_filters=512,
kernel_size=(3, 3),
padding='SAME',
stride=(1, 1),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv5_3_dw, conv5_3_pw])
conv5_4_dw, conv5_4_pw = depthwise_separable_conv2d(
'conv_ds_11',
conv5_3_pw,
width_multiplier=1,
num_filters=512,
kernel_size=(3, 3),
padding='SAME',
stride=(1, 1),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv5_4_dw, conv5_4_pw])
conv5_5_dw, conv5_5_pw = depthwise_separable_conv2d(
'conv_ds_12',
conv5_4_pw,
width_multiplier=1,
num_filters=512,
kernel_size=(3, 3),
padding='SAME',
stride=(1, 1),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv5_5_dw, conv5_5_pw])
conv5_6_dw, conv5_6_pw = depthwise_separable_conv2d(
'conv_ds_13',
conv5_5_pw,
width_multiplier=1,
num_filters=1024,
kernel_size=(3, 3),
padding='SAME',
stride=(2, 2),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv5_6_dw, conv5_6_pw])
############################################################################################
conv6_1_dw, conv6_1_pw = depthwise_separable_conv2d(
'conv_ds_14',
conv5_6_pw,
width_multiplier=1,
num_filters=1024,
kernel_size=(3, 3),
padding='SAME',
stride=(1, 1),
batchnorm_enabled=False,
activation=tf.nn.relu6,
is_training=True,
l2_strength=0.0,
biases=(0.0, 0.0))
#self.__add_to_nodes([conv6_1_dw, conv6_1_pw])
############################################################################################
avg_pool = avg_pool_2d(conv6_1_pw, size=(7, 7), stride=(1, 1))
dropped = dropout(avg_pool, -1, True)
#print("dropout:shape:")
#print(dropped.get_shape())
if 1:
logits = flatten(
conv2d('fc',
dropped,
kernel_size=(1, 1),
num_filters=32,
l2_strength=0.0,
bias=0.0,
padding='SAME'))
else:
logits = (conv2d('fc',
dropped,
kernel_size=(1, 1),
num_filters=32,
l2_strength=0.0,
bias=0.0,
padding='SAME'))
logits = tf_util.remove_axis(logits, [2, 3])
if 1:
logits_shape = logits.get_shape().as_list()
pool5_reshape = tf.reshape(
logits, [batch_size, num_unrolls, 2, logits_shape[-1]])
else:
skip_concat = tf.concat([conv1_skip_flat, logits], 1)
#skip_concat = tf.concat([conv1_skip_flat, conv2_skip_flat, logits], 1)
#
#print("logitss:shape:")
#print(logits_shape)
#
skip_concat_shape = skip_concat.get_shape().as_list()
print("Ship_concat shape")
print(skip_concat_shape)
# Split and merge image pairs
# (BxTx2)xHxWxC
pool5_reshape = tf.reshape(
skip_concat, [batch_size, num_unrolls, 2, skip_concat_shape[-1]])
# (BxT)x(2xHxWxC)
reshaped = tf_util.remove_axis(pool5_reshape, [1, 3])
return reshaped
def alexnet_conv_layers(input, batch_size, num_unrolls):
input = tf.to_float(input) - IMAGENET_MEAN
with tf.variable_scope('conv1'):
conv1 = tf_util.conv_layer(input, 96, 11, 4, padding='VALID')
pool1 = tf.nn.max_pool(conv1,
ksize=[1, 3, 3, 1],
strides=[1, 2, 2, 1],
padding='VALID',
name='pool1')
lrn1 = tf.nn.local_response_normalization(pool1,
depth_radius=2,
alpha=2e-5,
beta=0.75,
bias=1.0,
name='norm1')
with tf.variable_scope('conv1_skip'):
prelu_skip = tf_util.get_variable('prelu',
shape=[16],
dtype=tf.float32,
initializer=prelu_initializer)
conv1_skip = tf_util.prelu(
tf_util.conv_layer(lrn1, 16, 1, activation=None), prelu_skip)
conv1_skip = tf.transpose(conv1_skip, perm=[0, 3, 1, 2])
conv1_skip_flat = tf_util.remove_axis(conv1_skip, [2, 3])
with tf.variable_scope('conv2'):
conv2 = tf_util.conv_layer(lrn1, 256, 5, num_groups=2, padding='SAME')
pool2 = tf.nn.max_pool(conv2,
ksize=[1, 3, 3, 1],
strides=[1, 2, 2, 1],
padding='VALID',
name='pool2')
lrn2 = tf.nn.local_response_normalization(pool2,
depth_radius=2,
alpha=2e-5,
beta=0.75,
bias=1.0,
name='norm2')
with tf.variable_scope('conv2_skip'):
prelu_skip = tf_util.get_variable('prelu',
shape=[32],
dtype=tf.float32,
initializer=prelu_initializer)
conv2_skip = tf_util.prelu(
tf_util.conv_layer(lrn2, 32, 1, activation=None), prelu_skip)
conv2_skip = tf.transpose(conv2_skip, perm=[0, 3, 1, 2])
conv2_skip_flat = tf_util.remove_axis(conv2_skip, [2, 3])
with tf.variable_scope('conv3'):
conv3 = tf_util.conv_layer(lrn2, 384, 3, padding='SAME')
with tf.variable_scope('conv4'):
conv4 = tf_util.conv_layer(conv3, 384, 3, num_groups=2, padding='SAME')
with tf.variable_scope('conv5'):
conv5 = tf_util.conv_layer(conv4, 256, 3, num_groups=2, padding='SAME')
pool5 = tf.nn.max_pool(conv5,
ksize=[1, 3, 3, 1],
strides=[1, 2, 2, 1],
padding='VALID',
name='pool5')
pool5 = tf.transpose(pool5, perm=[0, 3, 1, 2])
pool5_flat = tf_util.remove_axis(pool5, [2, 3])
with tf.variable_scope('conv5_skip'):
prelu_skip = tf_util.get_variable('prelu',
shape=[64],
dtype=tf.float32,
initializer=prelu_initializer)
conv5_skip = tf_util.prelu(
tf_util.conv_layer(conv5, 64, 1, activation=None), prelu_skip)
conv5_skip = tf.transpose(conv5_skip, perm=[0, 3, 1, 2])
conv5_skip_flat = tf_util.remove_axis(conv5_skip, [2, 3])
with tf.variable_scope('big_concat'):
# Concat all skip layers.
skip_concat = tf.concat(
[conv1_skip_flat, conv2_skip_flat, conv5_skip_flat, pool5_flat], 1)
skip_concat_shape = skip_concat.get_shape().as_list()
# Split and merge image pairs
# (BxTx2)xHxWxC
pool5_reshape = tf.reshape(
skip_concat, [batch_size, num_unrolls, 2, skip_concat_shape[-1]])
# (BxT)x(2xHxWxC)
reshaped = tf_util.remove_axis(pool5_reshape, [1, 3])
return reshaped
def mobilenet_v2(model_file,
input,
batch_size,
num_unrolls,
image_size=128,
depth_multiplier=0.5):
from nets.mobilenet import mobilenet_v2
if model_file is None:
model_file = "/home/waechter/repos/tf-models/research/slim/nets/mobilenet/checkpoint/mobilenet_v2_" + str(
depth_multiplier) + "_" + str(image_size) + ".ckpt"
with tf.Session() as sess:
print("input: ", input.shape)
#image2 = tf.gather(input, [None,1,None,None, None])
#images = tf.expand_dims(input, 0)
# image1 = images[:,0,:,:,:] #tf.gather(input, [None,0,None,None, None])
# image2 = images[:,1,:,:,:] #tf.gather(input, [None,0,None,None, None])
# print("input: ", input.shape, " image: ", images.shape)
# print("input: ", input.shape, " image: ", images.shape, "image1: ", image1.shape)
images = tf.cast(input, tf.float32) / 128. - 1
print(images.shape)
images.set_shape((None, None, None, 3))
print("images shape after: ", input.shape)
images = tf.image.resize(images, (image_size, image_size))
with tf.contrib.slim.arg_scope(
mobilenet_v2.training_scope(is_training=True)):
# logits, endpoints = mobilenet_v2.mobilenet(images, depth_multiplier=depth_multiplier)
logits, endpoints = mobilenet_v2.mobilenet_v2_050(images)
ema = tf.train.ExponentialMovingAverage(0.999)
vars = ema.variables_to_restore()
changed_vars = {}
search_string = "MobilenetV2"
for key in vars.keys():
pos = int(key.rfind(search_string))
#print(key, "pos: ", pos, " str1:", key[pos:])
if (pos >= 0):
changed_vars[key[pos:]] = vars[key]
#print(zip(changed_vars.keys(), vars.keys()))
# print("Vars: ", vars)
saver = tf.train.Saver(changed_vars)
# saver.restore(sess, "/home/waechter/repos/tf-models/research/slim/nets/mobilenet/checkpoint/mobilenet_v2_1.0_192.ckpt")
saver.restore(sess, model_file)
print("Model mobilenet_v2 restored.")
conv_skip_connection1 = endpoints['layer_3']
conv_skip_connection2 = endpoints['layer_10']
print("skip1 shape: ", conv_skip_connection1.shape)
print("skip2 shape: ", conv_skip_connection2.shape)
final_conv = endpoints['layer_17']
print("layer_17 shape: ", conv_skip_connection1.shape)
final_conv_flat = tf_util.remove_axis(final_conv, [2, 3])
print("layer_17_flat shape: ", conv_skip_connection1.shape)
with tf.variable_scope('conv_skip1'):
prelu_skip = tf_util.get_variable('prelu',
shape=[16],
dtype=tf.float32,
initializer=prelu_initializer)
conv1_skip = tf_util.prelu(
tf_util.conv_layer(tf.stop_gradient(conv_skip_connection1),
16,
1,
activation=None), prelu_skip)
conv1_skip = tf.transpose(conv1_skip, perm=[0, 3, 1, 2])
conv1_skip_flat = tf_util.remove_axis(conv1_skip, [2, 3])
with tf.variable_scope('conv_skip2'):
prelu_skip = tf_util.get_variable('prelu',
shape=[16],
dtype=tf.float32,
initializer=prelu_initializer)
conv2_skip = tf_util.prelu(
tf_util.conv_layer(tf.stop_gradient(conv_skip_connection2),
16,
1,
activation=None), prelu_skip)
conv2_skip = tf.transpose(conv2_skip, perm=[0, 3, 1, 2])
conv2_skip_flat = tf_util.remove_axis(conv2_skip, [2, 3])
final_conv_flat = tf.stop_gradient(final_conv_flat)
with tf.variable_scope('big_concat'):
skip_concat = tf.concat(
[conv1_skip_flat, conv2_skip_flat, final_conv_flat], 1)
skip_concat_shape = skip_concat.get_shape().as_list()
print("skip_concat shape: ", skip_concat.shape)
# Split and merge image pairs
# (BxTx2)xHxWxC
concat_reshape = tf.reshape(
skip_concat, [batch_size, num_unrolls, 2, skip_concat_shape[-1]])
# (BxT)x(2xHxWxC)
reshaped = tf_util.remove_axis(concat_reshape, [1, 3])
return reshaped