def base(self, images, is_training):
use_bias = False
self.images = self.input = images
with tf.compat.v1.variable_scope("init"):
self.conv1 = conv2d(
"conv1",
self.images,
filters=16,
kernel_size=3,
activation=None,
use_bias=use_bias,
is_debug=self.is_debug,
)
self.bn1 = batch_norm("bn1", self.conv1, is_training=is_training)
with tf.compat.v1.variable_scope("relu1"):
self.relu1 = tf.nn.relu(self.bn1)
for i in range(0, self.num_residual):
with tf.compat.v1.variable_scope("unit1_{}".format(i)):
if i == 0:
out = self._residual(self.relu1, in_filters=16, out_filters=16, strides=1, is_training=is_training)
else:
out = self._residual(out, in_filters=16, out_filters=16, strides=1, is_training=is_training)
for i in range(0, self.num_residual):
with tf.compat.v1.variable_scope("unit2_{}".format(i)):
if i == 0:
out = self._residual(out, in_filters=16, out_filters=32, strides=2, is_training=is_training)
else:
out = self._residual(out, in_filters=32, out_filters=32, strides=1, is_training=is_training)
for i in range(0, self.num_residual):
with tf.compat.v1.variable_scope("unit3_{}".format(i)):
if i == 0:
out = self._residual(out, in_filters=32, out_filters=64, strides=2, is_training=is_training)
else:
out = self._residual(out, in_filters=64, out_filters=64, strides=1, is_training=is_training)
with tf.compat.v1.variable_scope("unit4_0"):
self.bn4 = batch_norm("bn4", out, is_training=is_training, activation=tf.nn.relu)
# global average pooling
h = self.bn4.get_shape()[1].value
w = self.bn4.get_shape()[2].value
self.global_average_pool = average_pooling2d(
"global_average_pool", self.bn4, pool_size=[h, w], padding="VALID", is_debug=self.is_debug,)
self._heatmap_layer = None
self.fc = fully_connected("fc", self.global_average_pool, filters=self.num_classes, activation=None)
return self.fc
def _residual(self, inputs, in_filters, out_filters, strides, is_training):
use_bias = False
with tf.compat.v1.variable_scope('sub1'):
bn1 = batch_norm("bn1", inputs, is_training=is_training)
with tf.compat.v1.variable_scope('relu1'):
relu1 = tf.nn.relu(bn1)
conv1 = conv2d(
"conv1",
relu1,
filters=out_filters,
kernel_size=3,
activation=None,
use_bias=use_bias,
strides=strides,
is_debug=self.is_debug,
)
with tf.compat.v1.variable_scope('sub2'):
bn2 = batch_norm("bn2", conv1, is_training=is_training)
with tf.compat.v1.variable_scope('relu2'):
relu2 = tf.nn.relu(bn2)
conv2 = conv2d(
"conv2",
relu2,
filters=out_filters,
kernel_size=3,
activation=None,
use_bias=use_bias,
strides=1,
is_debug=self.is_debug,
)
with tf.compat.v1.variable_scope('sub_add'):
if in_filters != out_filters:
inputs = tf.nn.avg_pool(
inputs,
ksize=[1, strides, strides, 1],
strides=[1, strides, strides, 1],
padding='VALID'
)
inputs = tf.pad(
inputs,
[[0, 0], [0, 0], [0, 0], [(out_filters - in_filters)//2, (out_filters - in_filters)//2]]
)
output = conv2 + inputs
return output
def base(self, images, is_training):
assert self.data_format == "NHWC"
channel_data_format = "channels_last"
self.inputs = self.images = images
with tf.compat.v1.variable_scope("block_1"):
conv = conv2d("conv", self.inputs, filters=32, kernel_size=3,
activation=None, use_bias=False, data_format=channel_data_format,
kernel_initializer=tf.compat.v1.keras.initializers.VarianceScaling(scale=2.0))
batch_normed = batch_norm("bn", conv, is_training=is_training, decay=0.99, scale=True, center=True,
data_format=self.data_format)
self.block_1 = self.activation(batch_normed)
self.block_last = conv2d("block_last", self.block_1, filters=self.num_classes, kernel_size=1,
activation=None, use_bias=True, is_debug=self.is_debug,
kernel_initializer=tf.compat.v1.random_normal_initializer(mean=0.0, stddev=0.01),
data_format=channel_data_format)
h = self.block_last.get_shape()[1].value
w = self.block_last.get_shape()[2].value
self.pool = tf.compat.v1.layers.average_pooling2d(name='global_average_pool', inputs=self.block_last,
pool_size=[h, w], padding='VALID', strides=1,
data_format=channel_data_format)
self.base_output = tf.reshape(self.pool, [-1, self.num_classes], name="pool_reshape")
return self.base_output
def darknet(name,
inputs,
filters,
kernel_size,
is_training=tf.constant(False),
activation=None,
data_format="NHWC"):
"""Darknet19 block.
Do convolution, batch_norm, bias, leaky_relu activation.
Ref: https://arxiv.org/pdf/1612.08242.pdf
https://github.com/pjreddie/darknet/blob/3bf2f342c03b0ad22efd799d5be9990c9d792354/cfg/darknet19.cfg
https://github.com/pjreddie/darknet/blob/8215a8864d4ad07e058acafd75b2c6ff6600b9e8/cfg/yolo.2.0.cfg
"""
if data_format == "NCHW":
channel_data_format = "channels_first"
elif data_format == "NHWC":
channel_data_format = "channels_last"
else:
raise ValueError(
"data format must be 'NCHW' or 'NHWC'. got {}.".format(
data_format))
with tf.compat.v1.variable_scope(name):
if activation is None:
def activation(x):
return tf.nn.leaky_relu(x, alpha=0.1, name="leaky_relu")
conv = conv2d(
"conv",
inputs,
filters=filters,
kernel_size=kernel_size,
activation=None,
use_bias=False,
data_format=channel_data_format,
kernel_initializer=tf.contrib.layers.variance_scaling_initializer(
),
) # he initializer
# TODO(wakisaka): Should be the same as darknet batch norm.
# https://github.com/tensorflow/tensorflow/blob/r1.1/tensorflow/contrib/layers/python/layers/layers.py
# https://github.com/pjreddie/darknet/blob/8215a8864d4ad07e058acafd75b2c6ff6600b9e8/src/batchnorm_layer.c#L135
batch_normed = batch_norm("bn",
conv,
is_training=is_training,
decay=0.99,
scale=True,
center=True,
data_format=data_format)
tf.compat.v1.summary.histogram("batch_normed", batch_normed)
output = activation(batch_normed)
tf.compat.v1.summary.histogram("output", output)
return output