def ResNet(stack_fn,
preact,
use_bias,
model_name='resnet',
include_top=True,
weights='imagenet',
input_tensor=None,
input_shape=None,
pooling=None,
classes=1000,
**kwargs):
"""Instantiates the ResNet, ResNetV2, and ResNeXt architecture.
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in your Keras config at `~/.keras/keras.json`.
Arguments:
stack_fn: a function that returns output tensor for the
stacked residual blocks.
preact: whether to use pre-activation or not
(True for ResNetV2, False for ResNet and ResNeXt).
use_bias: whether to use biases for convolutional layers or not
(True for ResNet and ResNetV2, False for ResNeXt).
model_name: string, model name.
include_top: whether to include the fully-connected
layer at the top of the network.
weights: one of `None` (random initialization),
'imagenet' (pre-training on ImageNet),
or the path to the weights file to be loaded.
input_tensor: optional Keras tensor
(i.e. output of `layers.Input()`)
to use as image input for the model.
input_shape: optional shape tuple, only to be specified
if `include_top` is False (otherwise the input shape
has to be `(224, 224, 3)` (with `channels_last` data format)
or `(3, 224, 224)` (with `channels_first` data format).
It should have exactly 3 inputs channels.
pooling: optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model will be
the 4D tensor output of the
last convolutional layer.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional layer, and thus
the output of the model will be a 2D tensor.
- `max` means that global max pooling will
be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is True, and
if no `weights` argument is specified.
**kwargs: For backwards compatibility only.
Returns:
A Keras model instance.
Raises:
ValueError: in case of invalid argument for `weights`,
or invalid input shape.
"""
if 'layers' in kwargs:
global layers
layers = kwargs.pop('layers')
if kwargs:
raise ValueError('Unknown argument(s): %s' % (kwargs,))
if not (weights in {'imagenet', None} or os.path.exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError('If using `weights` as `"imagenet"` with `include_top`'
' as true, `classes` should be 1000')
# Determine proper input shape
input_shape =obtain_input_shape(
input_shape,
default_size=224,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=include_top,
weights=weights)
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
bn_axis = 3 if backend.image_data_format() == 'channels_last' else 1
x = layers.ZeroPadding2D(
padding=((3, 3), (3, 3)), name='conv1_pad')(img_input)
x = layers.Conv2D(64, 7, strides=2, use_bias=use_bias, name='conv1_conv')(x)
if not preact:
x = layers.BatchNormalization(
axis=bn_axis, epsilon=1.001e-5, name='conv1_bn')(x)
x = layers.Activation('relu', name='conv1_relu')(x)
x = layers.ZeroPadding2D(padding=((1, 1), (1, 1)), name='pool1_pad')(x)
x = layers.MaxPooling2D(3, strides=2, name='pool1_pool')(x)
x = stack_fn(x)
if preact:
x = layers.BatchNormalization(
axis=bn_axis, epsilon=1.001e-5, name='post_bn')(x)
x = layers.Activation('relu', name='post_relu')(x)
if include_top:
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
x = layers.Dense(classes, activation='softmax', name='probs')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D(name='max_pool')(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
model = training.Model(inputs, x, name=model_name)
# Load weights.
if (weights == 'imagenet') and (model_name in WEIGHTS_HASHES):
if include_top:
file_name = model_name + '_weights_tf_dim_ordering_tf_kernels.h5'
file_hash = WEIGHTS_HASHES[model_name][0]
else:
file_name = model_name + '_weights_tf_dim_ordering_tf_kernels_notop.h5'
file_hash = WEIGHTS_HASHES[model_name][1]
weights_path = data_utils.get_file(
file_name,
BASE_WEIGHTS_PATH + file_name,
cache_subdir='models',
file_hash=file_hash)
model.load_weights(weights_path)
elif weights is not None:
model.load_weights(weights)
return model
def resnet50(num_classes, dtype='float32', batch_size=None):
"""Instantiates the ResNet50 architecture.
Args:
num_classes: `int` number of classes for image classification.
dtype: dtype to use float32 or float16 are most common.
batch_size: Size of the batches for each step.
Returns:
A Keras model instance.
"""
input_shape = (32, 64, 3)
img_input = layers.Input(shape=input_shape,
dtype=dtype,
batch_size=batch_size)
if backend.image_data_format() == 'channels_first':
x = layers.Lambda(
lambda x: backend.permute_dimensions(x, (0, 3, 1, 2)),
name='transpose')(img_input)
bn_axis = 1
else: # channels_last
x = img_input
bn_axis = 3
# x = layers.ZeroPadding2D(padding=(1, 1), name='conv1_pad')(x)
x = layers.Conv2D(16, (7, 7),
strides=(2, 2),
padding='same',
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name='conv1')(x)
x = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name='bn_conv1')(x)
x = layers.Activation('relu')(x)
x = layers.MaxPooling2D((2, 4), strides=(2, 4), padding='same')(x)
x = conv_block(x, 3, [16, 16, 64], stage=2, block='a', strides=(1, 1))
x = identity_block(x, 3, [16, 16, 64], stage=2, block='b')
x = identity_block(x, 3, [16, 16, 64], stage=2, block='c')
x = conv_block(x, 3, [32, 32, 128], stage=3, block='a')
x = identity_block(x, 3, [32, 32, 128], stage=3, block='b')
x = identity_block(x, 3, [32, 32, 128], stage=3, block='c')
x = identity_block(x, 3, [32, 32, 128], stage=3, block='d')
x = conv_block(x, 3, [64, 64, 256], stage=4, block='a')
x = identity_block(x, 3, [64, 64, 256], stage=4, block='b')
x = identity_block(x, 3, [64, 64, 256], stage=4, block='c')
x = identity_block(x, 3, [64, 64, 256], stage=4, block='d')
x = identity_block(x, 3, [64, 64, 256], stage=4, block='e')
x = identity_block(x, 3, [64, 64, 256], stage=4, block='f')
x = conv_block(x, 3, [128, 128, 512], stage=5, block='a')
x = identity_block(x, 3, [128, 128, 512], stage=5, block='b')
x = identity_block(x, 3, [128, 128, 512], stage=5, block='c')
rm_axes = [1, 2
] if backend.image_data_format() == 'channels_last' else [2, 3]
x = layers.Lambda(lambda x: backend.mean(x, rm_axes),
name='reduce_mean')(x)
x = layers.Dense(num_classes,
kernel_initializer=initializers.RandomNormal(stddev=0.01),
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
bias_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name='fc1000')(x)
# TODO(reedwm): Remove manual casts once mixed precision can be enabled with a
# single line of code.
x = backend.cast(x, 'float32')
x = layers.Activation('softmax')(x)
# Create model.
return models.Model(img_input, x, name='resnet50')
def __init__(self, num_classes=10, dtype="float32", batch_size=None):
super(CustomModel, self).__init__(name="resnet50")
if backend.image_data_format() == "channels_first":
self._lambda = layers.Lambda(
lambda x: backend.permute_dimensions(x, (0, 3, 1, 2)),
name="transpose",
)
bn_axis = 1
data_format = "channels_first"
else:
bn_axis = 3
data_format = "channels_last"
self._padding = layers.ZeroPadding2D(padding=(3, 3),
data_format=data_format,
name="zero_pad")
self._conv2d_1 = layers.Conv2D(
64,
(7, 7),
strides=(2, 2),
padding="valid",
use_bias=False,
kernel_initializer="he_normal",
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name="conv1",
)
self._bn_1 = layers.BatchNormalization(
axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name="bn_conv1",
)
self._activation_1 = layers.Activation("relu")
self._maxpooling2d = layers.MaxPooling2D((3, 3),
strides=(2, 2),
padding="same")
self._conv_block_1 = ConvBlock(3, [64, 64, 256],
stage=2,
block="a",
strides=(1, 1))
self._identity_block_1 = IdentityBlock(3, [64, 64, 256],
stage=2,
block="b")
self._identity_block_2 = IdentityBlock(3, [64, 64, 256],
stage=2,
block="c")
self._conv_block_2 = ConvBlock(3, [128, 128, 512], stage=3, block="a")
self._identity_block_3 = IdentityBlock(3, [128, 128, 512],
stage=3,
block="b")
self._identity_block_4 = IdentityBlock(3, [128, 128, 512],
stage=3,
block="c")
self._identity_block_5 = IdentityBlock(3, [128, 128, 512],
stage=3,
block="d")
self._conv_block_3 = ConvBlock(3, [256, 256, 1024], stage=4, block="a")
self._identity_block_6 = IdentityBlock(3, [256, 256, 1024],
stage=4,
block="b")
self._identity_block_7 = IdentityBlock(3, [256, 256, 1024],
stage=4,
block="c")
self._identity_block_8 = IdentityBlock(3, [256, 256, 1024],
stage=4,
block="d")
self._identity_block_9 = IdentityBlock(3, [256, 256, 1024],
stage=4,
block="e")
self._identity_block_10 = IdentityBlock(3, [256, 256, 1024],
stage=4,
block="f")
self._conv_block_4 = ConvBlock(3, [512, 512, 2048], stage=5, block="a")
self._identity_block_11 = IdentityBlock(3, [512, 512, 2048],
stage=5,
block="b")
self._identity_block_12 = IdentityBlock(3, [512, 512, 2048],
stage=5,
block="c")
rm_axes = ([1, 2] if backend.image_data_format() == "channels_last"
else [2, 3])
self._lamba_2 = layers.Lambda(lambda x: backend.mean(x, rm_axes),
name="reduce_mean")
self._dense = layers.Dense(
num_classes,
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
bias_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name="fc1000",
)
self._activation_2 = layers.Activation("softmax")
def __init__(self, out_channels, atrous_rate, norm_layer, norm_kwargs, conv_trainable=True):
super(ASPPConv, self).__init__()
self.conv = klayers.Conv2D(out_channels, kernel_size=3, padding='same', kernel_initializer='he_uniform',
dilation_rate=atrous_rate, use_bias=False, trainable=conv_trainable)
self.bn = norm_layer(**({} if norm_kwargs is None else norm_kwargs))
self.relu = klayers.ReLU()
def conv_building_block(input_tensor,
kernel_size,
filters,
stage,
block,
strides=(2, 2),
training=None):
"""A block that has a conv layer at shortcut.
Arguments:
input_tensor: input tensor
kernel_size: default 3, the kernel size of
middle conv layer at main path
filters: list of integers, the filters of 3 conv layer at main path
stage: integer, current stage label, used for generating layer names
block: current block label, used for generating layer names
strides: Strides for the first conv layer in the block.
training: Only used if training keras model with Estimator. In other
scenarios it is handled automatically.
Returns:
Output tensor for the block.
Note that from stage 3,
the first conv layer at main path is with strides=(2, 2)
And the shortcut should have strides=(2, 2) as well
"""
filters1, filters2 = filters
if tf.keras.backend.image_data_format() == 'channels_last':
bn_axis = 3
else:
bn_axis = 1
conv_name_base = 'res' + str(stage) + block + '_branch'
bn_name_base = 'bn' + str(stage) + block + '_branch'
x = layers.Conv2D(filters1,
kernel_size,
strides=strides,
padding='same',
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name=conv_name_base + '2a')(input_tensor)
x = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name=bn_name_base + '2a')(x,
training=training)
x = layers.Activation('relu')(x)
x = layers.Conv2D(filters2,
kernel_size,
padding='same',
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name=conv_name_base + '2b')(x)
x = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name=bn_name_base + '2b')(x,
training=training)
shortcut = layers.Conv2D(
filters2, (1, 1),
strides=strides,
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name=conv_name_base + '1')(input_tensor)
shortcut = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name=bn_name_base + '1')(
shortcut, training=training)
x = layers.add([x, shortcut])
x = layers.Activation('relu')(x)
return x
def _adjust_block(p, ip, filters, block_id=None):
"""Adjusts the input `previous path` to match the shape of the `input`.
Used in situations where the output number of filters needs to be changed.
Arguments:
p: Input tensor which needs to be modified
ip: Input tensor whose shape needs to be matched
filters: Number of output filters to be matched
block_id: String block_id
Returns:
Adjusted Keras tensor
"""
channel_dim = 1 if backend.image_data_format() == 'channels_first' else -1
img_dim = 2 if backend.image_data_format() == 'channels_first' else -2
ip_shape = backend.int_shape(ip)
if p is not None:
p_shape = backend.int_shape(p)
with backend.name_scope('adjust_block'):
if p is None:
p = ip
elif p_shape[img_dim] != ip_shape[img_dim]:
with backend.name_scope('adjust_reduction_block_%s' % block_id):
p = layers.Activation('relu', name='adjust_relu_1_%s' % block_id)(p)
p1 = layers.AveragePooling2D((1, 1),
strides=(2, 2),
padding='valid',
name='adjust_avg_pool_1_%s' % block_id)(
p)
p1 = layers.Conv2D(
filters // 2, (1, 1),
padding='same',
use_bias=False,
name='adjust_conv_1_%s' % block_id,
kernel_initializer='he_normal')(
p1)
p2 = layers.ZeroPadding2D(padding=((0, 1), (0, 1)))(p)
p2 = layers.Cropping2D(cropping=((1, 0), (1, 0)))(p2)
p2 = layers.AveragePooling2D((1, 1),
strides=(2, 2),
padding='valid',
name='adjust_avg_pool_2_%s' % block_id)(
p2)
p2 = layers.Conv2D(
filters // 2, (1, 1),
padding='same',
use_bias=False,
name='adjust_conv_2_%s' % block_id,
kernel_initializer='he_normal')(
p2)
p = layers.concatenate([p1, p2], axis=channel_dim)
p = layers.BatchNormalization(
axis=channel_dim,
momentum=0.9997,
epsilon=1e-3,
name='adjust_bn_%s' % block_id)(
p)
elif p_shape[channel_dim] != filters:
with backend.name_scope('adjust_projection_block_%s' % block_id):
p = layers.Activation('relu')(p)
p = layers.Conv2D(
filters, (1, 1),
strides=(1, 1),
padding='same',
name='adjust_conv_projection_%s' % block_id,
use_bias=False,
kernel_initializer='he_normal')(
p)
p = layers.BatchNormalization(
axis=channel_dim,
momentum=0.9997,
epsilon=1e-3,
name='adjust_bn_%s' % block_id)(
p)
return p
def NASNet(
input_shape=None,
penultimate_filters=4032,
num_blocks=6,
stem_block_filters=96,
skip_reduction=True,
filter_multiplier=2,
include_top=True,
weights=None,
input_tensor=None,
pooling=None,
classes=1000,
default_size=None,
classifier_activation='softmax',
):
"""Instantiates a NASNet model.
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in your Keras config at `~/.keras/keras.json`.
Arguments:
input_shape: Optional shape tuple, the input shape
is by default `(331, 331, 3)` for NASNetLarge and
`(224, 224, 3)` for NASNetMobile.
It should have exactly 3 input channels,
and width and height should be no smaller than 32.
E.g. `(224, 224, 3)` would be one valid value.
penultimate_filters: Number of filters in the penultimate layer.
NASNet models use the notation `NASNet (N @ P)`, where:
- N is the number of blocks
- P is the number of penultimate filters
num_blocks: Number of repeated blocks of the NASNet model.
NASNet models use the notation `NASNet (N @ P)`, where:
- N is the number of blocks
- P is the number of penultimate filters
stem_block_filters: Number of filters in the initial stem block
skip_reduction: Whether to skip the reduction step at the tail
end of the network.
filter_multiplier: Controls the width of the network.
- If `filter_multiplier` < 1.0, proportionally decreases the number
of filters in each layer.
- If `filter_multiplier` > 1.0, proportionally increases the number
of filters in each layer.
- If `filter_multiplier` = 1, default number of filters from the
paper are used at each layer.
include_top: Whether to include the fully-connected
layer at the top of the network.
weights: `None` (random initialization) or
`imagenet` (ImageNet weights)
input_tensor: Optional Keras tensor (i.e. output of
`layers.Input()`)
to use as image input for the model.
pooling: Optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model
will be the 4D tensor output of the
last convolutional block.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional block, and thus
the output of the model will be a
2D tensor.
- `max` means that global max pooling will
be applied.
classes: Optional number of classes to classify images
into, only to be specified if `include_top` is True, and
if no `weights` argument is specified.
default_size: Specifies the default image size of the model
classifier_activation: A `str` or callable. The activation function to use
on the "top" layer. Ignored unless `include_top=True`. Set
`classifier_activation=None` to return the logits of the "top" layer.
Returns:
A `keras.Model` instance.
Raises:
ValueError: In case of invalid argument for `weights`,
invalid input shape or invalid `penultimate_filters` value.
ValueError: if `classifier_activation` is not `softmax` or `None` when
using a pretrained top layer.
"""
if not (weights in {'imagenet', None} or os.path.exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError('If using `weights` as `"imagenet"` with `include_top` '
'as true, `classes` should be 1000')
if (isinstance(input_shape, tuple) and None in input_shape and
weights == 'imagenet'):
raise ValueError('When specifying the input shape of a NASNet'
' and loading `ImageNet` weights, '
'the input_shape argument must be static '
'(no None entries). Got: `input_shape=' +
str(input_shape) + '`.')
if default_size is None:
default_size = 331
# Determine proper input shape and default size.
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=default_size,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=True,
weights=weights)
if backend.image_data_format() != 'channels_last':
logging.warning('The NASNet family of models is only available '
'for the input data format "channels_last" '
'(width, height, channels). '
'However your settings specify the default '
'data format "channels_first" (channels, width, height).'
' You should set `image_data_format="channels_last"` '
'in your Keras config located at ~/.keras/keras.json. '
'The model being returned right now will expect inputs '
'to follow the "channels_last" data format.')
backend.set_image_data_format('channels_last')
old_data_format = 'channels_first'
else:
old_data_format = None
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
if penultimate_filters % (24 * (filter_multiplier**2)) != 0:
raise ValueError(
'For NASNet-A models, the `penultimate_filters` must be a multiple '
'of 24 * (`filter_multiplier` ** 2). Current value: %d' %
penultimate_filters)
channel_dim = 1 if backend.image_data_format() == 'channels_first' else -1
filters = penultimate_filters // 24
x = layers.Conv2D(
stem_block_filters, (3, 3),
strides=(2, 2),
padding='valid',
use_bias=False,
name='stem_conv1',
kernel_initializer='he_normal')(
img_input)
x = layers.BatchNormalization(
axis=channel_dim, momentum=0.9997, epsilon=1e-3, name='stem_bn1')(
x)
p = None
x, p = _reduction_a_cell(
x, p, filters // (filter_multiplier**2), block_id='stem_1')
x, p = _reduction_a_cell(
x, p, filters // filter_multiplier, block_id='stem_2')
for i in range(num_blocks):
x, p = _normal_a_cell(x, p, filters, block_id='%d' % (i))
x, p0 = _reduction_a_cell(
x, p, filters * filter_multiplier, block_id='reduce_%d' % (num_blocks))
p = p0 if not skip_reduction else p
for i in range(num_blocks):
x, p = _normal_a_cell(
x, p, filters * filter_multiplier, block_id='%d' % (num_blocks + i + 1))
x, p0 = _reduction_a_cell(
x,
p,
filters * filter_multiplier**2,
block_id='reduce_%d' % (2 * num_blocks))
p = p0 if not skip_reduction else p
for i in range(num_blocks):
x, p = _normal_a_cell(
x,
p,
filters * filter_multiplier**2,
block_id='%d' % (2 * num_blocks + i + 1))
x = layers.Activation('relu')(x)
if include_top:
x = layers.GlobalAveragePooling2D()(x)
imagenet_utils.validate_activation(classifier_activation, weights)
x = layers.Dense(classes, activation=classifier_activation,
name='predictions')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D()(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D()(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
model = training.Model(inputs, x, name='NASNet')
# Load weights.
if weights == 'imagenet':
if default_size == 224: # mobile version
if include_top:
weights_path = data_utils.get_file(
'nasnet_mobile.h5',
NASNET_MOBILE_WEIGHT_PATH,
cache_subdir='models',
file_hash='020fb642bf7360b370c678b08e0adf61')
else:
weights_path = data_utils.get_file(
'nasnet_mobile_no_top.h5',
NASNET_MOBILE_WEIGHT_PATH_NO_TOP,
cache_subdir='models',
file_hash='1ed92395b5b598bdda52abe5c0dbfd63')
model.load_weights(weights_path)
elif default_size == 331: # large version
if include_top:
weights_path = data_utils.get_file(
'nasnet_large.h5',
NASNET_LARGE_WEIGHT_PATH,
cache_subdir='models',
file_hash='11577c9a518f0070763c2b964a382f17')
else:
weights_path = data_utils.get_file(
'nasnet_large_no_top.h5',
NASNET_LARGE_WEIGHT_PATH_NO_TOP,
cache_subdir='models',
file_hash='d81d89dc07e6e56530c4e77faddd61b5')
model.load_weights(weights_path)
else:
raise ValueError('ImageNet weights can only be loaded with NASNetLarge'
' or NASNetMobile')
elif weights is not None:
model.load_weights(weights)
if old_data_format:
backend.set_image_data_format(old_data_format)
return model
def MobileNet(input_shape=None, alpha=1.0, depth_multiplier=1, dropout=1e-3):
if input_shape is None:
default_size = 224
else:
if backend.image_data_format() == 'channels_first':
rows = input_shape[1]
cols = input_shape[2]
else:
rows = input_shape[0]
cols = input_shape[1]
if rows == cols and rows in [128, 160, 192, 224]:
default_size = rows
else:
default_size = 224
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=default_size,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=True)
row_axis = 0 if backend.image_data_format() == 'channels_last' else 0
rows = input_shape[row_axis]
img_input = layers.Input(shape=input_shape)
x = _conv_block(img_input, 32, alpha, strides=(2, 2))
x = _depthwise_conv_block(x, 64, alpha, depth_multiplier, block_id=1)
x = _depthwise_conv_block(x,
128,
alpha,
depth_multiplier,
strides=(2, 2),
block_id=2)
x = _depthwise_conv_block(x, 128, alpha, depth_multiplier, block_id=3)
x = _depthwise_conv_block(x,
256,
alpha,
depth_multiplier,
strides=(2, 2),
block_id=4)
x = _depthwise_conv_block(x, 256, alpha, depth_multiplier, block_id=5)
x = _depthwise_conv_block(x,
512,
alpha,
depth_multiplier,
strides=(2, 2),
block_id=6)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=7)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=8)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=9)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=10)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=11)
x = _depthwise_conv_block(x,
1024,
alpha,
depth_multiplier,
strides=(2, 2),
block_id=12)
x = _depthwise_conv_block(x, 1024, alpha, depth_multiplier, block_id=13)
if backend.image_data_format() == 'channels_first':
shape = (int(1024 * alpha), 1, 1)
else:
shape = (1, 1, int(1024 * alpha))
x = layers.GlobalAveragePooling2D()(x)
x = layers.Reshape(shape, name='reshape_1')(x)
x = layers.Dropout(dropout, name='dropout')(x)
x = layers.Conv2D(NUM_CLASSES, (1, 1), padding='same',
name='conv_preds')(x)
x = layers.Reshape((NUM_CLASSES, ), name='reshape_2')(x)
imagenet_utils.validate_activation('softmax', None)
x = layers.Activation(activation='softmax', name='predictions')(x)
# Create model.
model = training.Model(img_input,
x,
name='mobilenet_%0.2f_%s' % (alpha, rows))
return model
def bottleneck_block(input_tensor,
filters,
kernel_size=3,
stride=1,
conv_shortcut=False,
name=None):
"""A residual block.
Arguments:
input_tensor: input tensor.
filters: integer, filters of the bottleneck layer.
kernel_size: default 3, kernel size of the bottleneck layer.
stride: default 1, stride of the first layer.
conv_shortcut: default False, use convolution shortcut if True,
otherwise identity shortcut.
name: string, block label.
Returns:
Output tensor for the residual block.
"""
if backend.image_data_format() == 'channels_last':
bn_axis = -1
else:
bn_axis = 1
preact = layers.BatchNormalization(axis=bn_axis,
name=name + '_preact_bn')(input_tensor)
preact = layers.Activation(activation='relu',
name=name + '_preact_relu')(preact)
if conv_shortcut:
shortcut = layers.Conv2D(filters=4 * filters,
kernel_size=1,
strides=stride,
name=name + '_0_conv')(preact)
else:
shortcut = input_tensor
x = layers.Conv2D(filters=filters,
kernel_size=1,
strides=1,
use_bias=False,
name=name + '_1_conv')(preact)
x = layers.BatchNormalization(axis=bn_axis, name=name + '_1_bn')(x)
x = layers.Activation(activation='relu', name=name + '_1_relu')(x)
x = layers.ZeroPadding2D(padding=1, name=name + '_2_pad')(x)
x = layers.Conv2D(filters=filters,
kernel_size=kernel_size,
strides=stride,
use_bias=False,
name=name + '_2_conv')(x)
x = layers.BatchNormalization(axis=bn_axis, name=name + '_2_bn')(x)
x = layers.Activation(activation='relu', name=name + '_2_relu')(x)
x = layers.Conv2D(filters=4 * filters,
kernel_size=1,
strides=1,
name=name + '_3_conv')(x)
x = layers.Add(name=name + '_out')([shortcut, x])
return x
def conv_block_test_0(input_tensor, num_filters):
encoder = layers.Conv2D(num_filters, (3, 3), padding='same')(input_tensor)
encoder = layers.Activation('relu')(encoder)
encoder = layers.Conv2D(num_filters, (3, 3), padding='same')(encoder)
encoder = layers.Activation('relu')(encoder)
return encoder
def testSupports_KerasLayer(self):
self.assertTrue(self.quantize_registry.supports(l.Dense(10)))
self.assertTrue(self.quantize_registry.supports(l.Conv2D(10, (2, 2))))
def __init__(self, num_classes=10):
super(GoogLeNet, self).__init__()
self.conv1 = Sequential([
layers.Conv2D(filters=64,
kernel_size=(7, 7),
strides=2,
padding='same'),
layers.MaxPool2D(pool_size=(3, 3), strides=2, padding='same'),
layers.Activation('relu')
])
self.conv2 = Sequential([
layers.Conv2D(filters=192,
kernel_size=(3, 3),
strides=1,
padding='same'),
layers.MaxPool2D(pool_size=(3, 3), strides=2, padding='same'),
layers.Activation('relu')
])
self.inc3a = Inception_Block(out_channels_1_1=64,
reduce_channels_3_3=96,
out_channels_3_3=128,
reduce_channels_5_5=16,
out_channels_5_5=32,
pool_proj=32,
check=256)
self.inc3b = Inception_Block(out_channels_1_1=128,
reduce_channels_3_3=128,
out_channels_3_3=192,
reduce_channels_5_5=32,
out_channels_5_5=96,
pool_proj=64,
check=480)
self.inc4a = Inception_Block(out_channels_1_1=192,
reduce_channels_3_3=96,
out_channels_3_3=208,
reduce_channels_5_5=16,
out_channels_5_5=48,
pool_proj=64,
check=512)
self.inc4b = Inception_Block(out_channels_1_1=160,
reduce_channels_3_3=112,
out_channels_3_3=224,
reduce_channels_5_5=24,
out_channels_5_5=64,
pool_proj=64,
check=512)
self.inc4c = Inception_Block(out_channels_1_1=128,
reduce_channels_3_3=128,
out_channels_3_3=256,
reduce_channels_5_5=24,
out_channels_5_5=64,
pool_proj=64,
check=512)
self.inc4d = Inception_Block(out_channels_1_1=112,
reduce_channels_3_3=144,
out_channels_3_3=288,
reduce_channels_5_5=32,
out_channels_5_5=64,
pool_proj=64,
check=528)
self.inc4e = Inception_Block(out_channels_1_1=256,
reduce_channels_3_3=160,
out_channels_3_3=320,
reduce_channels_5_5=32,
out_channels_5_5=128,
pool_proj=128,
check=832)
self.inc5a = Inception_Block(out_channels_1_1=256,
reduce_channels_3_3=160,
out_channels_3_3=320,
reduce_channels_5_5=32,
out_channels_5_5=128,
pool_proj=128,
check=832)
self.inc5b = Inception_Block(out_channels_1_1=384,
reduce_channels_3_3=192,
out_channels_3_3=384,
reduce_channels_5_5=48,
out_channels_5_5=128,
pool_proj=128,
check=1024)
self.last_seq = Sequential([
layers.GlobalAveragePooling2D(),
layers.Dropout(0.4),
layers.Dense(num_classes),
# layers.Dense(num_classes, activation=tf.nn.softmax)
])
def create_network(Zt, Ct,
filter_size=5,
channels=1,
strides=[2, 2, 2, 2, 2],
upscaling_filters=[512, 256, 128, 64, 1],
encoder_filters=[64, 128, 256],
decoder_filters=[256, 128, 64],
resblock_filters=[256, 256, 256, 256, 256],
resblock_ks=[3, 3, 3, 3, 3, 3]
):
with tf.name_scope("Gen"):
# Generator
Z = kl.Input((None, None, channels,), tensor=Zt, name="Z")
C = kl.Input((None, None, channels,), tensor=Ct, name="C")
layer = Z
# Upscaling
for l in range(len(upscaling_filters) - 1):
layer = kl.Conv2DTranspose(
filters=upscaling_filters[l],
kernel_size=filter_size,
padding="same",
strides=strides[l],
kernel_regularizer=kr.l2(),
activation="relu")(layer)
layer = kl.BatchNormalization()(layer)
layer = kl.Conv2DTranspose(
filters=upscaling_filters[-1],
kernel_size=filter_size,
padding="same",
strides=strides[-1],
activation="relu",
kernel_regularizer=kr.l2())(layer)
layer = kl.concatenate([layer, C])
# Encoder
skips = []
for l in range(len(encoder_filters)):
layer = kl.Conv2D(
filters=encoder_filters[l],
kernel_size=filter_size,
padding="same",
activation="relu",
kernel_regularizer=kr.l2())(layer)
layer = kl.AveragePooling2D()(layer)
layer = kl.BatchNormalization()(layer)
skips.append(layer)
# Residual blocks
for l in range(len(resblock_filters)):
layer = ResidualBlock(resblock_filters[l], nb_layers=3, kernel_size=resblock_ks[l])(layer)
# Decoder
skips = skips[::-1]
for l in range(len(decoder_filters)):
layer = kl.concatenate([layer, skips[l]])
layer = kl.Conv2DTranspose(
filters=decoder_filters[l],
kernel_size=filter_size,
padding="same",
strides=strides[l],
kernel_regularizer=kr.l2(),
activation="relu")(layer)
layer = kl.BatchNormalization()(layer)
G_out = kl.Conv2D(
filters=1,
kernel_size=filter_size,
activation="tanh",
padding="same",
kernel_regularizer=kr.l2())(layer)
model = k.Model(inputs=[Z, C], outputs=G_out, name="G")
return model
def init_model(learning_rate, dense_layers, dense_nodes, activation):
"""
Hyper-parameters:
learning_rate: Learning-rate for the optimizer.
dense_layers: Number of dense layers.
dense_nodes: Number of nodes in each dense layer.
activation: Activation function for all layers.
"""
# Start construction of a Keras Sequential model.
model = models.Sequential()
# Add an input layer which is similar to a feed_dict in TensorFlow.
# Note that the input-shape must be a tuple containing the image-size.
model.add(layers.InputLayer(input_shape=(IMG_SIZE_flat,)))
# The input from MNIST is a flattened array with 784 elements,
# but the convolutional layers expect images with shape (28, 28, 1)
model.add(layers.Reshape(IMG_SHAPE_FULL))
# First convolutional layer.
# There are many hyper-parameters in this layer, but we only
# want to optimize the activation-function in this example.
model.add(layers.Conv2D(
kernel_size=5, strides=1, filters=16, padding='same',
activation=activation, name='layer_conv1'))
model.add(layers.MaxPooling2D(pool_size=2, strides=2))
# Second convolutional layer.
# Again, we only want to optimize the activation-function here.
model.add(layers.Conv2D(
kernel_size=5, strides=1, filters=36, padding='same',
activation=activation, name='layer_conv2'))
model.add(layers.MaxPooling2D(pool_size=2, strides=2))
# Flatten the 4-rank output of the convolutional layers
# to 2-rank that can be input to a fully-connected / dense layer.
model.add(layers.Flatten())
# Add fully-connected / dense layers.
# The number of layers is a hyper-parameter we want to optimize.
for i in range(dense_layers):
# Name of the layer. This is not really necessary
# because Keras should give them unique names.
name = 'layer_dense_{0}'.format(i+1)
# Add the dense / fully-connected layer to the model.
# This has two hyper-parameters we want to optimize:
# The number of nodes and the activation function.
model.add(layers.Dense(
dense_nodes, activation=activation, name=name))
# Last fully-connected / dense layer with softmax-activation
# for use in classification.
model.add(layers.Dense(NUM_CLASSES, activation='softmax'))
# Use the Adam method for training the network.
# We want to find the best learning-rate for the Adam method.
optimizer = optimizers.Adam(lr=learning_rate)
# In Keras we need to compile the model so it can be trained.
model.compile(optimizer=optimizer,
loss='categorical_crossentropy',
metrics=['accuracy'])
return model
from tensorflow.python.keras import models, layers
from tensorflow.python.keras.utils import to_categorical
(train_images, train_labels), (test_images, test_labels) = mnist.load_data()
train_images = train_images.reshape((60000, 28, 28, 1))
train_images = train_images.astype('float32') / 255
test_images = test_images.reshape((10000, 28, 28, 1))
test_images = test_images.astype('float32') / 255
train_labels = to_categorical(train_labels)
test_labels = to_categorical(test_labels)
model = models.Sequential()
model.add(layers.Conv2D(32, (3, 3), activation='relu',
input_shape=(28, 28, 1)))
model.add(layers.MaxPool2D(2, 2))
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
model.add(layers.MaxPool2D(2, 2))
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
model.add(layers.Flatten())
model.add(layers.Dense(64, activation='relu'))
model.add(layers.Dense(10, activation='softmax'))
# print(model.summary())
model.compile(optimizer='rmsprop',
loss='categorical_crossentropy',
metrics=['accuracy'])
model.fit(train_images, train_labels, epochs=5, batch_size=64)
def conv_block(input_tensor, num_filters):
encoder = layers.Conv2D(num_filters, (3, 3), padding='same')(input_tensor)
encoder = layers.BatchNormalization()(encoder)
encoder = layers.LeakyReLU(alpha=0.3)(encoder)
return encoder
def DenseNet(blocks,
include_top=True,
weights='imagenet',
input_tensor=None,
input_shape=None,
pooling=None,
classes=1000,
**kwargs):
"""Instantiates the DenseNet architecture.
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in your Keras config at `~/.keras/keras.json`.
# Arguments
blocks: numbers of building blocks for the four dense layers.
include_top: whether to include the fully-connected
layer at the top of the network.
weights: one of `None` (random initialization),
'imagenet' (pre-training on ImageNet),
or the path to the weights file to be loaded.
input_tensor: optional Keras tensor
(i.e. output of `layers.Input()`)
to use as image input for the model.
input_shape: optional shape tuple, only to be specified
if `include_top` is False (otherwise the input shape
has to be `(224, 224, 3)` (with `'channels_last'` data format)
or `(3, 224, 224)` (with `'channels_first'` data format).
It should have exactly 3 inputs channels,
and width and height should be no smaller than 32.
E.g. `(200, 200, 3)` would be one valid value.
pooling: optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model will be
the 4D tensor output of the
last convolutional block.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional block, and thus
the output of the model will be a 2D tensor.
- `max` means that global max pooling will
be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is True, and
if no `weights` argument is specified.
# Returns
A Keras model instance.
# Raises
ValueError: in case of invalid argument for `weights`,
or invalid input shape.
"""
if not (weights in {'imagenet', None} or os.path.exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError(
'If using `weights` as `"imagenet"` with `include_top`'
' as true, `classes` should be 1000')
# Determine proper input shape
input_shape = _obtain_input_shape(input_shape,
default_size=224,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=include_top,
weights=weights)
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
bn_axis = 3 if backend.image_data_format() == 'channels_last' else 1
x = layers.ZeroPadding2D(padding=((3, 3), (3, 3)))(img_input)
x = layers.Conv2D(64, 7, strides=2, use_bias=False, name='conv1/conv')(x)
x = layers.BatchNormalization(axis=bn_axis,
epsilon=1.001e-5,
name='conv1/bn')(x)
x = layers.Activation('relu', name='conv1/relu')(x)
x = layers.ZeroPadding2D(padding=((1, 1), (1, 1)))(x)
x = layers.MaxPooling2D(3, strides=2, name='pool1')(x)
x = dense_block(x, blocks[0], name='conv2')
x = transition_block(x, 0.5, name='pool2')
x = dense_block(x, blocks[1], name='conv3')
x = transition_block(x, 0.5, name='pool3')
x = dense_block(x, blocks[2], name='conv4')
x = transition_block(x, 0.5, name='pool4')
x = dense_block(x, blocks[3], name='conv5')
x = layers.BatchNormalization(axis=bn_axis, epsilon=1.001e-5, name='bn')(x)
x = layers.Activation('relu', name='relu')(x)
if include_top:
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
x = layers.Dense(classes, activation='softmax', name='fc1000')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D(name='max_pool')(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = keras_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
if blocks == [6, 12, 24, 16]:
model = models.Model(inputs, x, name='densenet121')
else:
model = models.Model(inputs, x, name='densenet')
# Load weights.
if weights == 'imagenet':
if include_top or (blocks != [6, 12, 24, 16]):
raise ValueError('Only DenseNet121 without top is available'
' pretrained')
# NOTE: If your Hub Application has multiple configurations (e.g.,
# DenseNet {121, 169, 201} x {include_top=True, include_top=False})
# each of these would have their own weights_path/SavedModel.
weights_path = 'densenet121_weights_tf_dim_ordering_tf_kernels_notop'
# Download the SavedModel and use only the checkpoint (variables).
path = resolve(('https://github.com/jharmsen/hub-application/'
'releases/download/v1/{}.tar.gz').format(weights_path))
path = os.path.join(path, 'variables/variables')
model.load_weights(path)
elif weights is not None:
model.load_weights(weights)
return model
def RetinaNet(input_shape, num_classes, num_anchor=9):
inputs = tf.keras.Input(shape=input_shape)
# FPN
resnet50 = tf.keras.applications.ResNet50(weights="imagenet",
include_top=False,
input_tensor=inputs,
pooling=None)
assert resnet50.layers[80].name == "conv3_block4_out"
C3 = resnet50.layers[80].output
assert resnet50.layers[142].name == "conv4_block6_out"
C4 = resnet50.layers[142].output
assert resnet50.layers[-1].name == "conv5_block3_out"
C5 = resnet50.layers[-1].output
P5 = layers.Conv2D(256,
kernel_size=1,
strides=1,
padding='same',
kernel_regularizer=regularizers.l2(0.0001))(C5)
P5_upsampling = layers.UpSampling2D()(P5)
P4 = layers.Conv2D(256,
kernel_size=1,
strides=1,
padding='same',
kernel_regularizer=regularizers.l2(0.0001))(C4)
P4 = layers.Add()([P5_upsampling, P4])
P4_upsampling = layers.UpSampling2D()(P4)
P3 = layers.Conv2D(256,
kernel_size=1,
strides=1,
padding='same',
kernel_regularizer=regularizers.l2(0.0001))(C3)
P3 = layers.Add()([P4_upsampling, P3])
P6 = layers.Conv2D(256,
kernel_size=3,
strides=2,
padding='same',
name="P6",
kernel_regularizer=regularizers.l2(0.0001))(C5)
P7 = layers.Activation('relu')(P6)
P7 = layers.Conv2D(256,
kernel_size=3,
strides=2,
padding='same',
name="P7",
kernel_regularizer=regularizers.l2(0.0001))(P7)
P5 = layers.Conv2D(256,
kernel_size=3,
strides=1,
padding='same',
name="P5",
kernel_regularizer=regularizers.l2(0.0001))(P5)
P4 = layers.Conv2D(256,
kernel_size=3,
strides=1,
padding='same',
name="P4",
kernel_regularizer=regularizers.l2(0.0001))(P4)
P3 = layers.Conv2D(256,
kernel_size=3,
strides=1,
padding='same',
name="P3",
kernel_regularizer=regularizers.l2(0.0001))(P3)
# classification subnet
cls_subnet = _classification_sub_net(num_classes=num_classes,
num_anchor=num_anchor)
P3_cls = cls_subnet(P3)
P4_cls = cls_subnet(P4)
P5_cls = cls_subnet(P5)
P6_cls = cls_subnet(P6)
P7_cls = cls_subnet(P7)
cls_output = layers.Concatenate(axis=-2)(
[P3_cls, P4_cls, P5_cls, P6_cls, P7_cls])
# localization subnet
loc_subnet = _regression_sub_net(num_anchor=num_anchor)
P3_loc = loc_subnet(P3)
P4_loc = loc_subnet(P4)
P5_loc = loc_subnet(P5)
P6_loc = loc_subnet(P6)
P7_loc = loc_subnet(P7)
loc_output = layers.Concatenate(axis=-2)(
[P3_loc, P4_loc, P5_loc, P6_loc, P7_loc])
return tf.keras.Model(inputs=inputs, outputs=[cls_output, loc_output])
def _normal_a_cell(ip, p, filters, block_id=None):
"""Adds a Normal cell for NASNet-A (Fig. 4 in the paper).
Arguments:
ip: Input tensor `x`
p: Input tensor `p`
filters: Number of output filters
block_id: String block_id
Returns:
A Keras tensor
"""
channel_dim = 1 if backend.image_data_format() == 'channels_first' else -1
with backend.name_scope('normal_A_block_%s' % block_id):
p = _adjust_block(p, ip, filters, block_id)
h = layers.Activation('relu')(ip)
h = layers.Conv2D(
filters, (1, 1),
strides=(1, 1),
padding='same',
name='normal_conv_1_%s' % block_id,
use_bias=False,
kernel_initializer='he_normal')(
h)
h = layers.BatchNormalization(
axis=channel_dim,
momentum=0.9997,
epsilon=1e-3,
name='normal_bn_1_%s' % block_id)(
h)
with backend.name_scope('block_1'):
x1_1 = _separable_conv_block(
h, filters, kernel_size=(5, 5), block_id='normal_left1_%s' % block_id)
x1_2 = _separable_conv_block(
p, filters, block_id='normal_right1_%s' % block_id)
x1 = layers.add([x1_1, x1_2], name='normal_add_1_%s' % block_id)
with backend.name_scope('block_2'):
x2_1 = _separable_conv_block(
p, filters, (5, 5), block_id='normal_left2_%s' % block_id)
x2_2 = _separable_conv_block(
p, filters, (3, 3), block_id='normal_right2_%s' % block_id)
x2 = layers.add([x2_1, x2_2], name='normal_add_2_%s' % block_id)
with backend.name_scope('block_3'):
x3 = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same',
name='normal_left3_%s' % (block_id))(
h)
x3 = layers.add([x3, p], name='normal_add_3_%s' % block_id)
with backend.name_scope('block_4'):
x4_1 = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same',
name='normal_left4_%s' % (block_id))(
p)
x4_2 = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same',
name='normal_right4_%s' % (block_id))(
p)
x4 = layers.add([x4_1, x4_2], name='normal_add_4_%s' % block_id)
with backend.name_scope('block_5'):
x5 = _separable_conv_block(
h, filters, block_id='normal_left5_%s' % block_id)
x5 = layers.add([x5, h], name='normal_add_5_%s' % block_id)
x = layers.concatenate([p, x1, x2, x3, x4, x5],
axis=channel_dim,
name='normal_concat_%s' % block_id)
return x, ip
def structureModel(self):
Inputs = layers.Input(shape=self._inputShape, batch_size=self._iBatchSize)
Con1 = layers.Conv2D(64, (3, 3), name='Con1', activation='relu', padding='SAME', input_shape=self._inputShape, strides=1)(Inputs)
Con2 = layers.Conv2D(64, (3, 3), name='Con2', activation='relu', padding='SAME', strides=1)(Con1)
Side1 = sideBranch(Con2, 1)
MaxPooling1 = layers.MaxPooling2D((2, 2), name='MaxPooling1', strides=2, padding='SAME')(Con2)
# outputs1
Con3 = layers.Conv2D(128, (3, 3), name='Con3', activation='relu', padding='SAME', strides=1)(MaxPooling1)
Con4 = layers.Conv2D(128, (3, 3), name='Con4', activation='relu', padding='SAME', strides=1)(Con3)
Side2 = sideBranch(Con4, 2)
MaxPooling2 = layers.MaxPooling2D((2, 2), name='MaxPooling2', strides=2, padding='SAME')(Con4)
# outputs2
Con5 = layers.Conv2D(256, (3, 3), name='Con5', activation='relu', padding='SAME', strides=1)(MaxPooling2)
Con6 = layers.Conv2D(256, (3, 3), name='Con6', activation='relu', padding='SAME', strides=1)(Con5)
Con7 = layers.Conv2D(256, (3, 3), name='Con7', activation='relu', padding='SAME', strides=1)(Con6)
Side3 = sideBranch(Con7, 4)
MaxPooling3 = layers.MaxPooling2D((2, 2), name='MaxPooling3', strides=2, padding='SAME')(Con7)
# outputs3
Con8 = layers.Conv2D(512, (3, 3), name='Con8', activation='relu', padding='SAME', strides=1)(MaxPooling3)
Con9 = layers.Conv2D(512, (3, 3), name='Con9', activation='relu', padding='SAME', strides=1)(Con8)
Con10 = layers.Conv2D(512, (3, 3), name='Con10', activation='relu', padding='SAME', strides=1)(Con9)
Side4 = sideBranch(Con10, 8)
MaxPooling4 = layers.MaxPooling2D((2, 2), name='MaxPooling4', strides=2, padding='SAME')(Con10)
# outputs4
Con11 = layers.Conv2D(512, (3, 3), name='Con11', activation='relu', padding='SAME', strides=1)(MaxPooling4)
Con12 = layers.Conv2D(512, (3, 3), name='Con12', activation='relu', padding='SAME', strides=1)(Con11)
Con13 = layers.Conv2D(512, (3, 3), name='Con13', activation='relu', padding='SAME', strides=1)(Con12)
Side5 = sideBranch(Con13, 16)
Fuse = layers.Concatenate(axis=-1)([Side1, Side2, Side3, Side4, Side5])
# learn fusion weight
Fuse = layers.Conv2D(1, (1, 1), name='Fuse', padding='SAME', use_bias=False, activation=None)(Fuse)
output1 = layers.Activation('sigmoid', name='output1')(Side1)
output2 = layers.Activation('sigmoid', name='output2')(Side2)
output3 = layers.Activation('sigmoid', name='output3')(Side3)
output4 = layers.Activation('sigmoid', name='output4')(Side4)
output5 = layers.Activation('sigmoid', name='output5')(Side5)
output6 = layers.Activation('sigmoid', name='output6')(Fuse)
outputs = [output1, output2, output3, output4, output5, output6]
self._pModel = Model(inputs=Inputs, outputs=outputs)
pAdam = optimizers.adam(lr=0.0001)
self._pModel.compile(loss={'output1': classBalancedSigmoidCrossEntropy,
'output2': classBalancedSigmoidCrossEntropy,
'output3': classBalancedSigmoidCrossEntropy,
'output4': classBalancedSigmoidCrossEntropy,
'output5': classBalancedSigmoidCrossEntropy,
'output6': classBalancedSigmoidCrossEntropy
}, optimizer=pAdam)
def _reduction_a_cell(ip, p, filters, block_id=None):
"""Adds a Reduction cell for NASNet-A (Fig. 4 in the paper).
Arguments:
ip: Input tensor `x`
p: Input tensor `p`
filters: Number of output filters
block_id: String block_id
Returns:
A Keras tensor
"""
channel_dim = 1 if backend.image_data_format() == 'channels_first' else -1
with backend.name_scope('reduction_A_block_%s' % block_id):
p = _adjust_block(p, ip, filters, block_id)
h = layers.Activation('relu')(ip)
h = layers.Conv2D(
filters, (1, 1),
strides=(1, 1),
padding='same',
name='reduction_conv_1_%s' % block_id,
use_bias=False,
kernel_initializer='he_normal')(
h)
h = layers.BatchNormalization(
axis=channel_dim,
momentum=0.9997,
epsilon=1e-3,
name='reduction_bn_1_%s' % block_id)(
h)
h3 = layers.ZeroPadding2D(
padding=imagenet_utils.correct_pad(h, 3),
name='reduction_pad_1_%s' % block_id)(
h)
with backend.name_scope('block_1'):
x1_1 = _separable_conv_block(
h,
filters, (5, 5),
strides=(2, 2),
block_id='reduction_left1_%s' % block_id)
x1_2 = _separable_conv_block(
p,
filters, (7, 7),
strides=(2, 2),
block_id='reduction_right1_%s' % block_id)
x1 = layers.add([x1_1, x1_2], name='reduction_add_1_%s' % block_id)
with backend.name_scope('block_2'):
x2_1 = layers.MaxPooling2D((3, 3),
strides=(2, 2),
padding='valid',
name='reduction_left2_%s' % block_id)(
h3)
x2_2 = _separable_conv_block(
p,
filters, (7, 7),
strides=(2, 2),
block_id='reduction_right2_%s' % block_id)
x2 = layers.add([x2_1, x2_2], name='reduction_add_2_%s' % block_id)
with backend.name_scope('block_3'):
x3_1 = layers.AveragePooling2D((3, 3),
strides=(2, 2),
padding='valid',
name='reduction_left3_%s' % block_id)(
h3)
x3_2 = _separable_conv_block(
p,
filters, (5, 5),
strides=(2, 2),
block_id='reduction_right3_%s' % block_id)
x3 = layers.add([x3_1, x3_2], name='reduction_add3_%s' % block_id)
with backend.name_scope('block_4'):
x4 = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same',
name='reduction_left4_%s' % block_id)(
x1)
x4 = layers.add([x2, x4])
with backend.name_scope('block_5'):
x5_1 = _separable_conv_block(
x1, filters, (3, 3), block_id='reduction_left4_%s' % block_id)
x5_2 = layers.MaxPooling2D((3, 3),
strides=(2, 2),
padding='valid',
name='reduction_right5_%s' % block_id)(
h3)
x5 = layers.add([x5_1, x5_2], name='reduction_add4_%s' % block_id)
x = layers.concatenate([x2, x3, x4, x5],
axis=channel_dim,
name='reduction_concat_%s' % block_id)
return x, ip
def MobileNetV2(input_shape=None, alpha=1.0):
if input_shape is None:
default_size = 224
else:
if backend.image_data_format() == 'channels_first':
rows = input_shape[1]
cols = input_shape[2]
else:
rows = input_shape[0]
cols = input_shape[1]
if rows == cols and rows in [96, 128, 160, 192, 224]:
default_size = rows
else:
default_size = 224
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=default_size,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=True)
row_axis = 0 if backend.image_data_format() == 'channels_last' else 1
rows = input_shape[row_axis]
img_input = layers.Input(shape=input_shape)
channel_axis = 1 if backend.image_data_format() == 'channels_first' else -1
first_block_filters = _make_divisible(32 * alpha, 8)
x = layers.Conv2D(first_block_filters,
kernel_size=3,
strides=(2, 2),
padding='same',
use_bias=False,
name='Conv1')(img_input)
x = layers.BatchNormalization(axis=channel_axis,
epsilon=1e-3,
momentum=0.999,
name='bn_Conv1')(x)
x = layers.ReLU(6., name='Conv1_relu')(x)
x = _inverted_res_block(x,
filters=16,
alpha=alpha,
stride=1,
expansion=1,
block_id=0)
x = _inverted_res_block(x,
filters=24,
alpha=alpha,
stride=2,
expansion=6,
block_id=1)
x = _inverted_res_block(x,
filters=24,
alpha=alpha,
stride=1,
expansion=6,
block_id=2)
x = _inverted_res_block(x,
filters=32,
alpha=alpha,
stride=2,
expansion=6,
block_id=3)
x = _inverted_res_block(x,
filters=32,
alpha=alpha,
stride=1,
expansion=6,
block_id=4)
x = _inverted_res_block(x,
filters=32,
alpha=alpha,
stride=1,
expansion=6,
block_id=5)
x = _inverted_res_block(x,
filters=64,
alpha=alpha,
stride=2,
expansion=6,
block_id=6)
x = _inverted_res_block(x,
filters=64,
alpha=alpha,
stride=1,
expansion=6,
block_id=7)
x = _inverted_res_block(x,
filters=64,
alpha=alpha,
stride=1,
expansion=6,
block_id=8)
x = _inverted_res_block(x,
filters=64,
alpha=alpha,
stride=1,
expansion=6,
block_id=9)
x = _inverted_res_block(x,
filters=96,
alpha=alpha,
stride=1,
expansion=6,
block_id=10)
x = _inverted_res_block(x,
filters=96,
alpha=alpha,
stride=1,
expansion=6,
block_id=11)
x = _inverted_res_block(x,
filters=96,
alpha=alpha,
stride=1,
expansion=6,
block_id=12)
x = _inverted_res_block(x,
filters=160,
alpha=alpha,
stride=2,
expansion=6,
block_id=13)
x = _inverted_res_block(x,
filters=160,
alpha=alpha,
stride=1,
expansion=6,
block_id=14)
x = _inverted_res_block(x,
filters=160,
alpha=alpha,
stride=1,
expansion=6,
block_id=15)
x = _inverted_res_block(x,
filters=320,
alpha=alpha,
stride=1,
expansion=6,
block_id=16)
# no alpha applied to last conv as stated in the paper:
# if the width multiplier is greater than 1 we
# increase the number of output channels
if alpha > 1.0:
last_block_filters = _make_divisible(1280 * alpha, 8)
else:
last_block_filters = 1280
x = layers.Conv2D(last_block_filters,
kernel_size=1,
use_bias=False,
name='Conv_1')(x)
x = layers.BatchNormalization(axis=channel_axis,
epsilon=1e-3,
momentum=0.999,
name='Conv_1_bn')(x)
x = layers.ReLU(6., name='out_relu')(x)
x = layers.GlobalAveragePooling2D()(x)
imagenet_utils.validate_activation('softmax', None)
x = layers.Dense(NUM_CLASSES, activation='softmax', name='predictions')(x)
# Create model.
model = training.Model(img_input,
x,
name='mobilenetv2_%0.2f_%s' % (alpha, rows))
return model
def resnet(num_blocks, classes=10, training=None):
"""Instantiates the ResNet architecture.
Arguments:
num_blocks: integer, the number of conv/identity blocks in each block.
The ResNet contains 3 blocks with each block containing one conv block
followed by (layers_per_block - 1) number of idenity blocks. Each
conv/idenity block has 2 convolutional layers. With the input
convolutional layer and the pooling layer towards the end, this brings
the total size of the network to (6*num_blocks + 2)
classes: optional number of classes to classify images into
training: Only used if training keras model with Estimator. In other
scenarios it is handled automatically.
Returns:
A Keras model instance.
"""
input_shape = (32, 32, 3)
img_input = layers.Input(shape=input_shape)
if backend.image_data_format() == 'channels_first':
x = layers.Lambda(
lambda x: backend.permute_dimensions(x, (0, 3, 1, 2)),
name='transpose')(img_input)
bn_axis = 1
else: # channel_last
x = img_input
bn_axis = 3
x = layers.ZeroPadding2D(padding=(1, 1), name='conv1_pad')(x)
x = layers.Conv2D(16, (3, 3),
strides=(1, 1),
padding='valid',
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name='conv1')(x)
x = layers.BatchNormalization(
axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name='bn_conv1',
)(x, training=training)
x = layers.Activation('relu')(x)
x = resnet_block(x,
size=num_blocks,
kernel_size=3,
filters=[16, 16],
stage=2,
conv_strides=(1, 1),
training=training)
x = resnet_block(x,
size=num_blocks,
kernel_size=3,
filters=[32, 32],
stage=3,
conv_strides=(2, 2),
training=training)
x = resnet_block(x,
size=num_blocks,
kernel_size=3,
filters=[64, 64],
stage=4,
conv_strides=(2, 2),
training=training)
rm_axes = [1, 2
] if backend.image_data_format() == 'channels_last' else [2, 3]
x = layers.Lambda(lambda x: backend.mean(x, rm_axes),
name='reduce_mean')(x)
x = layers.Dense(classes,
activation='softmax',
kernel_initializer=initializers.RandomNormal(stddev=0.01),
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
bias_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name='fc10')(x)
inputs = img_input
# Create model.
model = tf.keras.models.Model(inputs, x, name='resnet56')
return model
def DenseNet(
blocks,
include_top=True,
weights='imagenet',
input_tensor=None,
input_shape=None,
pooling=None,
classes=1000,
classifier_activation='softmax',
):
"""Instantiates the DenseNet architecture.
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in your Keras config at `~/.keras/keras.json`.
Caution: Be sure to properly pre-process your inputs to the application.
Please see `applications.densenet.preprocess_input` for an example.
Arguments:
blocks: numbers of building blocks for the four dense layers.
include_top: whether to include the fully-connected
layer at the top of the network.
weights: one of `None` (random initialization),
'imagenet' (pre-training on ImageNet),
or the path to the weights file to be loaded.
input_tensor: optional Keras tensor
(i.e. output of `layers.Input()`)
to use as image input for the model.
input_shape: optional shape tuple, only to be specified
if `include_top` is False (otherwise the input shape
has to be `(224, 224, 3)` (with `'channels_last'` data format)
or `(3, 224, 224)` (with `'channels_first'` data format).
It should have exactly 3 inputs channels,
and width and height should be no smaller than 32.
E.g. `(200, 200, 3)` would be one valid value.
pooling: optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model will be
the 4D tensor output of the
last convolutional block.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional block, and thus
the output of the model will be a 2D tensor.
- `max` means that global max pooling will
be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is True, and
if no `weights` argument is specified.
classifier_activation: A `str` or callable. The activation function to use
on the "top" layer. Ignored unless `include_top=True`. Set
`classifier_activation=None` to return the logits of the "top" layer.
Returns:
A `keras.Model` instance.
Raises:
ValueError: in case of invalid argument for `weights`,
or invalid input shape.
ValueError: if `classifier_activation` is not `softmax` or `None` when
using a pretrained top layer.
"""
if not (weights in {'imagenet', None} or os.path.exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError(
'If using `weights` as `"imagenet"` with `include_top`'
' as true, `classes` should be 1000')
# Determine proper input shape
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=224,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=include_top,
weights=weights)
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
bn_axis = 3 if backend.image_data_format() == 'channels_last' else 1
x = layers.ZeroPadding2D(padding=((3, 3), (3, 3)))(img_input)
x = layers.Conv2D(64, 7, strides=2, use_bias=False, name='conv1/conv')(x)
x = layers.BatchNormalization(axis=bn_axis,
epsilon=1.001e-5,
name='conv1/bn')(x)
x = layers.Activation('relu', name='conv1/relu')(x)
x = layers.ZeroPadding2D(padding=((1, 1), (1, 1)))(x)
x = layers.MaxPooling2D(3, strides=2, name='pool1')(x)
x = dense_block(x, blocks[0], name='conv2')
x = transition_block(x, 0.5, name='pool2')
x = dense_block(x, blocks[1], name='conv3')
x = transition_block(x, 0.5, name='pool3')
x = dense_block(x, blocks[2], name='conv4')
x = transition_block(x, 0.5, name='pool4')
x = dense_block(x, blocks[3], name='conv5')
x = layers.BatchNormalization(axis=bn_axis, epsilon=1.001e-5, name='bn')(x)
x = layers.Activation('relu', name='relu')(x)
if include_top:
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
imagenet_utils.validate_activation(classifier_activation, weights)
x = layers.Dense(classes,
activation=classifier_activation,
name='predictions')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D(name='max_pool')(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
if blocks == [6, 12, 24, 16]:
model = training.Model(inputs, x, name='densenet121')
elif blocks == [6, 12, 32, 32]:
model = training.Model(inputs, x, name='densenet169')
elif blocks == [6, 12, 48, 32]:
model = training.Model(inputs, x, name='densenet201')
else:
model = training.Model(inputs, x, name='densenet')
# Load weights.
if weights == 'imagenet':
if include_top:
if blocks == [6, 12, 24, 16]:
weights_path = data_utils.get_file(
'densenet121_weights_tf_dim_ordering_tf_kernels.h5',
DENSENET121_WEIGHT_PATH,
cache_subdir='models',
file_hash='9d60b8095a5708f2dcce2bca79d332c7')
elif blocks == [6, 12, 32, 32]:
weights_path = data_utils.get_file(
'densenet169_weights_tf_dim_ordering_tf_kernels.h5',
DENSENET169_WEIGHT_PATH,
cache_subdir='models',
file_hash='d699b8f76981ab1b30698df4c175e90b')
elif blocks == [6, 12, 48, 32]:
weights_path = data_utils.get_file(
'densenet201_weights_tf_dim_ordering_tf_kernels.h5',
DENSENET201_WEIGHT_PATH,
cache_subdir='models',
file_hash='1ceb130c1ea1b78c3bf6114dbdfd8807')
else:
if blocks == [6, 12, 24, 16]:
weights_path = data_utils.get_file(
'densenet121_weights_tf_dim_ordering_tf_kernels_notop.h5',
DENSENET121_WEIGHT_PATH_NO_TOP,
cache_subdir='models',
file_hash='30ee3e1110167f948a6b9946edeeb738')
elif blocks == [6, 12, 32, 32]:
weights_path = data_utils.get_file(
'densenet169_weights_tf_dim_ordering_tf_kernels_notop.h5',
DENSENET169_WEIGHT_PATH_NO_TOP,
cache_subdir='models',
file_hash='b8c4d4c20dd625c148057b9ff1c1176b')
elif blocks == [6, 12, 48, 32]:
weights_path = data_utils.get_file(
'densenet201_weights_tf_dim_ordering_tf_kernels_notop.h5',
DENSENET201_WEIGHT_PATH_NO_TOP,
cache_subdir='models',
file_hash='c13680b51ded0fb44dff2d8f86ac8bb1')
model.load_weights(weights_path)
elif weights is not None:
model.load_weights(weights)
return model
def get_cnn_model():
# cnn_model.add(layers.Conv2D(
# filters=64,
# strides=[2, 2],
# kernel_size=[2, 2],
# padding="same",
# activation=activations.relu,
# input_shape=input_shape
# ))
# cnn_model.add(layers.MaxPooling2D(pool_size=[2, 2], strides=2, padding="same"))
#
# cnn_model.add(layers.Conv2D(
# filters=32,
# strides=[2, 2],
# kernel_size=[2, 2],
# padding="same",
# activation=activations.relu
# ))
# cnn_model.add(layers.MaxPooling2D(pool_size=[2, 2], strides=2, padding="same"))
# cnn_model.add(layers.Dropout(0.25))
# cnn_model.add(layers.Flatten())
# cnn_model.add(layers.Dropout(0.25))
# cnn_model.add(layers.Dense(
# units=128,
# activation=activations.relu,
# ))
# cnn_model.add(layers.Dropout(0.25))
# cnn_model.add(layers.Dense(num_classes, tf.nn.softmax))
#
# cnn_model.compile(
# optimizer=optimizers.Adagrad(),
# loss=losses.categorical_crossentropy,
# metrics=[metrics.top_k_categorical_accuracy]
# )
input = layers.Input(shape=input_shape, dtype=tf.float32, name="x")
c1 = layers.Conv2D(filters=32,
kernel_size=[2, 2],
strides=2,
padding="same",
activation=activations.relu)(input)
mp1 = layers.MaxPooling2D(strides=2, padding="same", pool_size=[2, 2])(c1)
c2 = layers.Conv2D(filters=64,
kernel_size=[2, 2],
strides=2,
padding="same",
activation=activations.relu)(mp1)
mp2 = layers.MaxPooling2D(strides=2, padding="same", pool_size=[2, 2])(c2)
flat = layers.Flatten()(mp2)
dense1 = layers.Dense(units=1024, activation=activations.relu)(flat)
output = layers.Dense(units=num_classes,
activation=activations.softmax)(dense1)
cnn_model = models.Model(inputs=input, outputs=output)
cnn_model.compile(optimizer=optimizers.Adagrad(),
loss=losses.categorical_crossentropy,
metrics=[
metrics.categorical_accuracy,
metrics.binary_accuracy, metrics.mean_squared_error
])
return cnn_model
x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train /= 255
x_test /= 255
y_train = keras.utils.to_categorical(y_train, num_classes)
y_test = keras.utils.to_categorical(y_test, num_classes)
##############################
# Create model
##############################
model = models.Sequential()
for cnn_depth in range(1, 3):
model.add(
layers.Conv2D(32 * cnn_depth,
kernel_size=(3, 3),
strides=(1, 1),
activation='relu',
input_shape=input_shape,
kernel_regularizer=regularizers.l2(l2_weights)))
model.add(layers.Activation("relu"))
model.add(layers.BatchNormalization())
model.add(
layers.Conv2D(32 * cnn_depth,
kernel_size=(3, 3),
strides=(1, 1),
activation='relu',
input_shape=input_shape,
kernel_regularizer=regularizers.l2(l2_weights)))
model.add(layers.Activation("relu"))
model.add(layers.BatchNormalization())
model.add(layers.MaxPooling2D(pool_size=(2, 2), strides=(2, 2)))
model.add(layers.Dropout(rate=0.1 + 0.025 * cnn_depth))
def conv_block(input_tensor,
kernel_size,
filters,
stage,
block,
strides=(2, 2)):
"""A block that has a conv layer at shortcut.
Note that from stage 3,
the second conv layer at main path is with strides=(2, 2)
And the shortcut should have strides=(2, 2) as well
Args:
input_tensor: input tensor
kernel_size: default 3, the kernel size of middle conv layer at main path
filters: list of integers, the filters of 3 conv layer at main path
stage: integer, current stage label, used for generating layer names
block: 'a','b'..., current block label, used for generating layer names
strides: Strides for the second conv layer in the block.
Returns:
Output tensor for the block.
"""
filters1, filters2, filters3 = filters
if backend.image_data_format() == 'channels_last':
bn_axis = 3
else:
bn_axis = 1
conv_name_base = 'res' + str(stage) + block + '_branch'
bn_name_base = 'bn' + str(stage) + block + '_branch'
x = layers.Conv2D(filters1, (1, 1),
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name=conv_name_base + '2a')(input_tensor)
x = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name=bn_name_base + '2a')(x)
x = layers.Activation('relu')(x)
x = layers.Conv2D(filters2,
kernel_size,
strides=strides,
padding='same',
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name=conv_name_base + '2b')(x)
x = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name=bn_name_base + '2b')(x)
x = layers.Activation('relu')(x)
x = layers.Conv2D(filters3, (1, 1),
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name=conv_name_base + '2c')(x)
x = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name=bn_name_base + '2c')(x)
shortcut = layers.Conv2D(
filters3, (1, 1),
strides=strides,
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name=conv_name_base + '1')(input_tensor)
shortcut = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name=bn_name_base + '1')(shortcut)
x = layers.add([x, shortcut])
x = layers.Activation('relu')(x)
return x
def define_model(self) -> keras.Model:
x = Input(shape=[
self.model_parameters.img_height, self.model_parameters.img_width,
self.model_parameters.num_channels
])
z = Input(shape=[
self.model_parameters.img_height, self.model_parameters.img_width,
self.model_parameters.num_channels
])
xz = z
if self.model_parameters.has_input_images:
xz += x
xz = layers.Conv2D(
filters=32,
kernel_size=(3, 3),
padding='same',
use_bias=False,
)(xz)
xz = layers.BatchNormalization()(xz)
xz = layers.LeakyReLU(alpha=0.2)(xz)
xz = layers.Conv2D(
filters=32,
kernel_size=(3, 3),
padding='same',
use_bias=False,
)(xz)
xz = layers.BatchNormalization()(xz)
xz = layers.LeakyReLU(alpha=0.2)(xz)
xz = layers.Conv2D(
filters=32,
kernel_size=(3, 3),
padding='same',
use_bias=False,
)(xz)
xz = layers.BatchNormalization()(xz)
xz = layers.LeakyReLU(alpha=0.2)(xz)
xz = layers.Conv2D(
filters=32,
kernel_size=(3, 3),
padding='same',
use_bias=False,
)(xz)
xz = layers.BatchNormalization()(xz)
xz = layers.LeakyReLU(alpha=0.2)(xz)
xz = layers.Conv2D(
filters=3,
kernel_size=(3, 3),
padding='same',
activation='tanh',
use_bias=False,
)(xz)
if self.model_parameters.has_input_images:
xz += x
model = Model(name=self.model_name, inputs=[x, z], outputs=xz)
return model
decoder4 = decoder_block(center, encoder4, 512)
# 16
decoder3 = decoder_block(decoder4, encoder3, 256)
# 32
decoder2 = decoder_block(decoder3, encoder2, 128)
# 64
decoder1 = decoder_block(decoder2, encoder1, 64)
# 128
decoder0 = decoder_block(decoder1, encoder0, 32)
# 256
outputs = layers.Conv2D(1, (1, 1), activation='sigmoid')(decoder0)
model = models.Model(inputs=[inputs], outputs=[outputs])
def dice_coeff(y_true, y_pred):
smooth = 1.
# Flatten
y_true_f = tf.reshape(y_true, [-1])
y_pred_f = tf.reshape(y_pred, [-1])
intersection = tf.reduce_sum(y_true_f * y_pred_f)
score = (2. * intersection + smooth) / (tf.reduce_sum(y_true_f) +
tf.reduce_sum(y_pred_f) + smooth)
return score
def block3(x,
filters,
kernel_size=3,
stride=1,
groups=32,
conv_shortcut=True,
name=None):
"""A residual block.
Arguments:
x: input tensor.
filters: integer, filters of the bottleneck layer.
kernel_size: default 3, kernel size of the bottleneck layer.
stride: default 1, stride of the first layer.
groups: default 32, group size for grouped convolution.
conv_shortcut: default True, use convolution shortcut if True,
otherwise identity shortcut.
name: string, block label.
Returns:
Output tensor for the residual block.
"""
bn_axis = 3 if backend.image_data_format() == 'channels_last' else 1
if conv_shortcut:
shortcut = layers.Conv2D(
(64 // groups) * filters,
1,
strides=stride,
use_bias=False,
name=name + '_0_conv')(x)
shortcut = layers.BatchNormalization(
axis=bn_axis, epsilon=1.001e-5, name=name + '_0_bn')(shortcut)
else:
shortcut = x
x = layers.Conv2D(filters, 1, use_bias=False, name=name + '_1_conv')(x)
x = layers.BatchNormalization(
axis=bn_axis, epsilon=1.001e-5, name=name + '_1_bn')(x)
x = layers.Activation('relu', name=name + '_1_relu')(x)
c = filters // groups
x = layers.ZeroPadding2D(padding=((1, 1), (1, 1)), name=name + '_2_pad')(x)
x = layers.DepthwiseConv2D(
kernel_size,
strides=stride,
depth_multiplier=c,
use_bias=False,
name=name + '_2_conv')(x)
x_shape = backend.int_shape(x)[1:-1]
x = layers.Reshape(x_shape + (groups, c, c))(x)
x = layers.Lambda(
lambda x: sum(x[:, :, :, :, i] for i in range(c)),
name=name + '_2_reduce')(x)
x = layers.Reshape(x_shape + (filters,))(x)
x = layers.BatchNormalization(
axis=bn_axis, epsilon=1.001e-5, name=name + '_2_bn')(x)
x = layers.Activation('relu', name=name + '_2_relu')(x)
x = layers.Conv2D(
(64 // groups) * filters, 1, use_bias=False, name=name + '_3_conv')(x)
x = layers.BatchNormalization(
axis=bn_axis, epsilon=1.001e-5, name=name + '_3_bn')(x)
x = layers.Add(name=name + '_add')([shortcut, x])
x = layers.Activation('relu', name=name + '_out')(x)
return x
