def _block_group(self,
inputs: tf.Tensor,
in_filters: int,
out_filters: int,
strides: int,
expand_ratio: int = 6,
block_repeats: int = 1,
se_ratio: float = 0.2,
stochastic_depth_drop_rate: Optional[float] = None,
name: str = 'block_group'):
"""Creates one group of blocks for the SpineNet model."""
x = nn_blocks.InvertedBottleneckBlock(
in_filters=in_filters,
out_filters=out_filters,
strides=strides,
se_ratio=se_ratio,
expand_ratio=expand_ratio,
stochastic_depth_drop_rate=stochastic_depth_drop_rate,
kernel_initializer=self._kernel_initializer,
kernel_regularizer=self._kernel_regularizer,
bias_regularizer=self._bias_regularizer,
activation=self._activation,
use_sync_bn=self._use_sync_bn,
norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon)(
inputs)
for _ in range(1, block_repeats):
x = nn_blocks.InvertedBottleneckBlock(
in_filters=in_filters,
out_filters=out_filters,
strides=1,
se_ratio=se_ratio,
expand_ratio=expand_ratio,
stochastic_depth_drop_rate=stochastic_depth_drop_rate,
kernel_initializer=self._kernel_initializer,
kernel_regularizer=self._kernel_regularizer,
bias_regularizer=self._bias_regularizer,
activation=self._activation,
use_sync_bn=self._use_sync_bn,
norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon)(
inputs)
return tf.identity(x, name=name)
def _mobilenet_base(
self,
inputs: tf.Tensor) -> Tuple[tf.Tensor, Dict[str, tf.Tensor], int]:
"""Builds the base MobileNet architecture.
Args:
inputs: A `tf.Tensor` of shape `[batch_size, height, width, channels]`.
Returns:
A tuple of output Tensor and dictionary that collects endpoints.
"""
input_shape = inputs.get_shape().as_list()
if len(input_shape) != 4:
raise ValueError('Expected rank 4 input, was: %d' %
len(input_shape))
# The current_stride variable keeps track of the output stride of the
# activations, i.e., the running product of convolution strides up to the
# current network layer. This allows us to invoke atrous convolution
# whenever applying the next convolution would result in the activations
# having output stride larger than the target output_stride.
current_stride = 1
# The atrous convolution rate parameter.
rate = 1
net = inputs
endpoints = {}
endpoint_level = 2
for i, block_def in enumerate(self._decoded_specs):
block_name = 'block_group_{}_{}'.format(block_def.block_fn, i)
# A small catch for gpooling block with None strides
if not block_def.strides:
block_def.strides = 1
if (self._output_stride is not None
and current_stride == self._output_stride):
# If we have reached the target output_stride, then we need to employ
# atrous convolution with stride=1 and multiply the atrous rate by the
# current unit's stride for use in subsequent layers.
layer_stride = 1
layer_rate = rate
rate *= block_def.strides
else:
layer_stride = block_def.strides
layer_rate = 1
current_stride *= block_def.strides
if block_def.block_fn == 'convbn':
net = Conv2DBNBlock(
filters=block_def.filters,
kernel_size=block_def.kernel_size,
strides=block_def.strides,
activation=block_def.activation,
use_bias=block_def.use_bias,
use_normalization=block_def.use_normalization,
kernel_initializer=self._kernel_initializer,
kernel_regularizer=self._kernel_regularizer,
bias_regularizer=self._bias_regularizer,
use_sync_bn=self._use_sync_bn,
norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon)(net)
elif block_def.block_fn == 'depsepconv':
net = nn_blocks.DepthwiseSeparableConvBlock(
filters=block_def.filters,
kernel_size=block_def.kernel_size,
strides=block_def.strides,
activation=block_def.activation,
dilation_rate=layer_rate,
regularize_depthwise=self._regularize_depthwise,
kernel_initializer=self._kernel_initializer,
kernel_regularizer=self._kernel_regularizer,
use_sync_bn=self._use_sync_bn,
norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon,
)(net)
elif block_def.block_fn == 'invertedbottleneck':
use_rate = rate
if layer_rate > 1 and block_def.kernel_size != 1:
# We will apply atrous rate in the following cases:
# 1) When kernel_size is not in params, the operation then uses
# default kernel size 3x3.
# 2) When kernel_size is in params, and if the kernel_size is not
# equal to (1, 1) (there is no need to apply atrous convolution to
# any 1x1 convolution).
use_rate = layer_rate
in_filters = net.shape.as_list()[-1]
net = nn_blocks.InvertedBottleneckBlock(
in_filters=in_filters,
out_filters=block_def.filters,
kernel_size=block_def.kernel_size,
strides=layer_stride,
expand_ratio=block_def.expand_ratio,
se_ratio=block_def.se_ratio,
expand_se_in_filters=True,
se_gating_activation='hard_sigmoid',
activation=block_def.activation,
use_depthwise=block_def.use_depthwise,
use_residual=block_def.use_residual,
dilation_rate=use_rate,
regularize_depthwise=self._regularize_depthwise,
kernel_initializer=self._kernel_initializer,
kernel_regularizer=self._kernel_regularizer,
bias_regularizer=self._bias_regularizer,
use_sync_bn=self._use_sync_bn,
norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon,
stochastic_depth_drop_rate=self.
_stochastic_depth_drop_rate,
divisible_by=self._get_divisible_by())(net)
elif block_def.block_fn == 'gpooling':
net = layers.GlobalAveragePooling2D()(net)
net = layers.Reshape((1, 1, net.shape[1]))(net)
else:
raise ValueError('Unknown block type {} for layer {}'.format(
block_def.block_fn, i))
net = tf.identity(net, name=block_name)
if block_def.is_output:
endpoints[str(endpoint_level)] = net
endpoint_level += 1
return net, endpoints, endpoint_level
def _mobiledet_base(self,
inputs: tf.Tensor
) -> Tuple[tf.Tensor, Dict[str, tf.Tensor], int]:
"""Builds the base MobileDet architecture.
Args:
inputs: A `tf.Tensor` of shape `[batch_size, height, width, channels]`.
Returns:
A tuple of output Tensor and dictionary that collects endpoints.
"""
input_shape = inputs.get_shape().as_list()
if len(input_shape) != 4:
raise ValueError('Expected rank 4 input, was: %d' % len(input_shape))
net = inputs
endpoints = {}
endpoint_level = 1
for i, block_def in enumerate(self._decoded_specs):
block_name = 'block_group_{}_{}'.format(block_def.block_fn, i)
if block_def.block_fn == 'convbn':
net = mobilenet.Conv2DBNBlock(
filters=block_def.filters,
kernel_size=block_def.kernel_size,
strides=block_def.strides,
activation=block_def.activation,
use_bias=block_def.use_bias,
use_normalization=block_def.use_normalization,
kernel_initializer=self._kernel_initializer,
kernel_regularizer=self._kernel_regularizer,
bias_regularizer=self._bias_regularizer,
use_sync_bn=self._use_sync_bn,
norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon
)(net)
elif block_def.block_fn == 'invertedbottleneck':
in_filters = net.shape.as_list()[-1]
net = nn_blocks.InvertedBottleneckBlock(
in_filters=in_filters,
out_filters=block_def.filters,
kernel_size=block_def.kernel_size,
strides=block_def.strides,
expand_ratio=block_def.expand_ratio,
se_ratio=block_def.se_ratio,
se_inner_activation=block_def.activation,
se_gating_activation='sigmoid',
se_round_down_protect=False,
expand_se_in_filters=True,
activation=block_def.activation,
use_depthwise=block_def.use_depthwise,
use_residual=block_def.use_residual,
regularize_depthwise=self._regularize_depthwise,
kernel_initializer=self._kernel_initializer,
kernel_regularizer=self._kernel_regularizer,
bias_regularizer=self._bias_regularizer,
use_sync_bn=self._use_sync_bn,
norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon,
divisible_by=self._get_divisible_by()
)(net)
elif block_def.block_fn == 'tucker':
in_filters = net.shape.as_list()[-1]
net = nn_blocks.TuckerConvBlock(
in_filters=in_filters,
out_filters=block_def.filters,
kernel_size=block_def.kernel_size,
strides=block_def.strides,
input_compression_ratio=block_def.input_compression_ratio,
output_compression_ratio=block_def.output_compression_ratio,
activation=block_def.activation,
use_residual=block_def.use_residual,
kernel_initializer=self._kernel_initializer,
kernel_regularizer=self._kernel_regularizer,
bias_regularizer=self._bias_regularizer,
use_sync_bn=self._use_sync_bn,
norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon,
divisible_by=self._get_divisible_by()
)(net)
else:
raise ValueError('Unknown block type {} for layer {}'.format(
block_def.block_fn, i))
net = tf.keras.layers.Activation('linear', name=block_name)(net)
if block_def.is_output:
endpoints[str(endpoint_level)] = net
endpoint_level += 1
return net, endpoints, endpoint_level
