def mobilenetv2_base(img_batch, is_training=True):
with slim.arg_scope(mobilenetv2_scope(is_training=is_training, trainable=True)):
feature_to_crop, endpoints = mobilenet_v2.mobilenet_base(input_tensor=img_batch,
num_classes=None,
is_training=False,
depth_multiplier=1.0,
scope='MobilenetV2',
conv_defs=V2_BASE_DEF,
finegrain_classification_mode=False)
# feature_to_crop = tf.Print(feature_to_crop, [tf.shape(feature_to_crop)], summarize=10, message='rpn_shape')
return feature_to_crop
def mobilenetv2_head(inputs, is_training=True):
with slim.arg_scope(
mobilenetv2_scope(is_training=is_training, trainable=True)):
net, _ = mobilenet_v2.mobilenet_base(
input_tensor=inputs,
num_classes=None,
is_training=False,
depth_multiplier=1.0,
scope='MobilenetV2',
conv_defs=V2_HEAD_DEF,
finegrain_classification_mode=True)
# net = tf.squeeze(net, [1, 2])
return net
def mobilenetv2_base(img_batch, is_training=True):
with slim.arg_scope(
mobilenetv2_scope(is_training=is_training, trainable=True)):
feature_to_crop, endpoints = mobilenet_v2.mobilenet_base(
input_tensor=img_batch,
num_classes=None,
is_training=False,
depth_multiplier=1.0,
scope='MobilenetV2',
conv_defs=V2_BASE_DEF,
finegrain_classification_mode=False)
feature_dict = {
"C3": endpoints['layer_5'],
'C4': endpoints['layer_8'],
'C5': endpoints['layer_15']
}
pyramid_dict = {}
with tf.variable_scope('build_pyramid'):
with slim.arg_scope([slim.conv2d],
weights_regularizer=slim.l2_regularizer(
cfgs.WEIGHT_DECAY),
activation_fn=None,
normalizer_fn=None):
P5 = slim.conv2d(feature_dict['C5'],
num_outputs=256,
kernel_size=[1, 1],
stride=1,
scope='build_P5')
pyramid_dict['P5'] = P5
for level in range(4, 2, -1): # build [P4, P3]
pyramid_dict['P%d' % level] = fusion_two_layer(
C_i=feature_dict["C%d" % level],
P_j=pyramid_dict["P%d" % (level + 1)],
scope='build_P%d' % level)
for level in range(5, 2, -1):
pyramid_dict['P%d' % level] = slim.conv2d(
pyramid_dict['P%d' % level],
num_outputs=256,
kernel_size=[3, 3],
padding="SAME",
stride=1,
scope="fuse_P%d" % level)
p6 = slim.conv2d(
pyramid_dict['P5'] if cfgs.USE_P5 else feature_dict['C5'],
num_outputs=256,
kernel_size=[3, 3],
padding="SAME",
stride=2,
scope='p6_conv')
pyramid_dict['P6'] = p6
p7 = tf.nn.relu(p6, name='p6_relu')
p7 = slim.conv2d(p7,
num_outputs=256,
kernel_size=[3, 3],
padding="SAME",
stride=2,
scope='p7_conv')
pyramid_dict['P7'] = p7
# for level in range(7, 1, -1):
# add_heatmap(pyramid_dict['P%d' % level], name='Layer%d/P%d_heat' % (level, level))
return pyramid_dict
def mobilenetv2_base(img_batch, is_training=True):
with slim.arg_scope(
mobilenetv2_scope(is_training=is_training, trainable=True)):
feature_to_crop, endpoints = mobilenet_v2.mobilenet_base(
input_tensor=img_batch,
num_classes=None,
is_training=False,
depth_multiplier=1.0,
scope='MobilenetV2',
conv_defs=V2_BASE_DEF,
finegrain_classification_mode=False)
feature_dict = {
"C3": endpoints['layer_5'],
'C4': endpoints['layer_8'],
'C5': endpoints['layer_15']
}
pyramid_dict = {}
with tf.variable_scope('build_pyramid'):
with slim.arg_scope([slim.conv2d],
weights_regularizer=slim.l2_regularizer(
cfgs.WEIGHT_DECAY),
activation_fn=None,
normalizer_fn=None):
P5 = slim.conv2d(feature_dict['C5'],
num_outputs=256,
kernel_size=[1, 1],
stride=1,
scope='build_P5')
pyramid_dict['P5'] = P5
for level in range(4, 2, -1): # build [P4, P3]
pyramid_dict['P%d' % level] = fusion_two_layer(
C_i=feature_dict["C%d" % level],
P_j=pyramid_dict["P%d" % (level + 1)],
scope='build_P%d' % level)
for level in range(5, 2, -1):
pyramid_dict['P%d' % level] = slim.conv2d(
pyramid_dict['P%d' % level],
num_outputs=256,
kernel_size=[3, 3],
padding="SAME",
stride=1,
scope="fuse_P%d" % level)
p6 = slim.conv2d(
pyramid_dict['P5'] if cfgs.USE_P5 else feature_dict['C5'],
num_outputs=256,
kernel_size=[3, 3],
padding="SAME",
stride=2,
scope='p6_conv')
pyramid_dict['P6'] = p6
p7 = tf.nn.relu(p6, name='p6_relu')
p7 = slim.conv2d(p7,
num_outputs=256,
kernel_size=[3, 3],
padding="SAME",
stride=2,
scope='p7_conv')
pyramid_dict['P7'] = p7
print("we are in Pyramid::-======>>>>")
print(cfgs.LEVEL)
print("base_anchor_size are: ", cfgs.BASE_ANCHOR_SIZE_LIST)
print(20 * "__")
if cfgs.USE_SUPERVISED_MASK:
mask_list = []
dot_layer_list = []
with tf.variable_scope("enrich_semantics"):
with slim.arg_scope([slim.conv2d],
weights_regularizer=slim.l2_regularizer(
cfgs.WEIGHT_DECAY),
normalizer_fn=None):
for i, l_name in enumerate(cfgs.GENERATE_MASK_LIST):
G, mask, dot_layer = generate_mask(
net=pyramid_dict[l_name],
num_layer=cfgs.ADDITION_LAYERS[i],
level_name=l_name)
# add_heatmap(G, name="MASK/G_%s" % l_name)
# add_heatmap(mask, name="MASK/mask_%s" % l_name)
# if cfgs.MASK_ACT_FET:
# pyramid_dict[l_name] = pyramid_dict[l_name] * dot_layer
dot_layer_list.append(dot_layer)
mask_list.append(mask)
return pyramid_dict, mask_list, dot_layer_list
else:
return pyramid_dict