def res_block(self,
input_,
output_dim,
is_downsizing=True,
name="res_block"):
with tf.variable_scope(name):
act1 = self.res_act(input_, name='act1')
if is_downsizing:
skip2 = layers.bottleneck_layer(act1,
output_dim,
d_h=2,
d_w=2,
name='skip1')
_, conv1 = layers.conv2d_same_act(
act1,
output_dim,
d_h=2,
d_w=2,
activation_fn=self.activation_fn,
with_logit=True,
name='conv1')
else:
skip2 = layers.bottleneck_layer(act1,
output_dim,
d_h=1,
d_w=1,
name='skip1')
_, conv1 = layers.conv2d_same_act(
act1,
output_dim,
d_h=1,
d_w=1,
activation_fn=self.activation_fn,
with_logit=True,
name='conv1')
conv2 = layers.conv2d_same(conv1, output_dim, name='conv2')
res1 = tf.add(skip2, conv2, name='res1')
act2 = self.res_act(res1, name='act2')
_, conv3 = layers.conv2d_same_repeat(
act2,
output_dim,
num_repeat=2,
d_h=1,
d_w=1,
activation_fn=self.activation_fn,
with_logit=True,
name='conv3')
res2 = tf.add(res1, conv3, name='res2')
return res2
def conv_res_conv_block(self,
input_,
output_dim,
name="conv_res_conv_block"):
with tf.variable_scope(name):
conv1 = layers.conv2d_same_act(input_,
output_dim,
activation_fn=self.act_fn,
with_logit=False,
name="conv1")
res = self.res_block_with_3_conv_layers(conv1,
output_dim,
name="res_block")
conv2 = layers.conv2d_same_act(res,
output_dim,
activation_fn=self.act_fn,
with_logit=False,
name="conv2")
return conv2
def basicBlock(self, input_, output_dim, stride=1, name='basic'):
assert output_dim is not 0
with tf.variable_scope(name):
input_dim = input_.get_shape().as_list()[-1]
bn = layers.batch_norm(input_, name='bn')
act = self.act_fn(bn, name='act')
conv1 = layers.conv2d_same_act(act, input_dim, activation_fn=self.act_fn, name='conv_1')
conv2 = layers.conv2d_same(conv1, output_dim, d_h=stride, d_w=stride, name='conv_2')
# make sure that identity is input_ and output of this block is conv2
sc = self.shortcut(conv2, input_, name='short')
return sc
def inference(self, input_, reuse=False):
with tf.variable_scope('ResNet') as scope:
if reuse:
scope.reuse_variables()
conv1 = layers.conv2d_same_act(input_, self.num_kernel, k_h=7, k_w=7, d_h=2, d_w=2,
activation_fn=self.act_fn, name='conv_1')
pool1 = layers.max_pool(conv1, k_h=self.pool_kernel, k_w=self.pool_kernel,
padding='SAME', name='pool1')
layer_blocks = self.layer_repeat(pool1, self.layer_def, name='layers')
pool2 = layers.global_avg_pool(layer_blocks, name='pool2')
flat = layers.flatten(pool2, 'flat')
linear = layers.linear(flat, self.num_class, name='linear')
logit = tf.sigmoid(linear, name='logit')
return logit
def _inference(self, input_):
conv1 = layers.conv2d_same_act(input_,
16,
activation_fn=self.activation_fn,
name='conv1')
skip1 = layers.bottleneck_layer(conv1, 32, name='skip1')
_, conv2 = layers.conv2d_same_repeat(conv1,
32,
num_repeat=2,
activation_fn=self.activation_fn,
with_logit=True,
name='conv2')
res1 = tf.add(skip1, conv2, name='res1')
res_act1 = self.res_act(res1)
_, conv3 = layers.conv2d_same_repeat(res_act1,
32,
num_repeat=2,
activation_fn=self.activation_fn,
with_logit=True,
name='conv3')
res2 = tf.add(conv3, res1, name='res2')
res_act2 = self.res_act(res2)
skip2 = layers.bottleneck_layer(res_act2,
64,
d_h=2,
d_w=2,
name='skip2')
conv4 = layers.conv2d_same_act(res_act2,
64,
d_h=2,
d_w=2,
activation_fn=self.activation_fn,
name='conv4')
conv5 = layers.conv2d_same(conv4, 64, name='conv5')
res3 = tf.add(skip2, conv5, name='res3')
res_act3 = self.res_act(res3)
_, conv6 = layers.conv2d_same_repeat(res_act1,
64,
num_repeat=2,
activation_fn=self.activation_fn,
with_logit=True,
name='conv3')
res4 = tf.add(res3, conv6, name='res4')
res_act4 = self.res_act(res4)
skip3 = layers.bottleneck_layer(res_act4,
128,
d_h=2,
d_w=2,
name='skip3')
conv7 = layers.conv2d_same_act(res_act4,
128,
d_h=2,
d_w=2,
activation_fn=self.activation_fn,
name='conv7')
conv8 = layers.conv2d_same(conv7, 128, name='conv8')
res5 = tf.add(skip3, conv8, name='res5')
res_act5 = self.res_act(res5)
_, conv9 = layers.conv2d_same_repeat(res_act5,
128,
num_repeat=2,
activation_fn=self.activation_fn,
with_logit=True,
name='conv9')
res6 = tf.add(res5, conv9, name='res6')
res_act6 = self.res_act(res6)
pool = layers.avg_pool(res_act6,
k_h=8,
k_w=8,
d_h=1,
d_w=1,
name='pool')
flat = layers.flatten(pool, 'flat')
linear = layers.linear(flat, self.num_class, name='linear')
return linear