def rpn_cls_net(self, inputs, scope_list, reuse_flag, level):
rpn_conv2d_3x3 = inputs
for i in range(4):
rpn_conv2d_3x3 = slim.conv2d(inputs=rpn_conv2d_3x3,
num_outputs=256,
kernel_size=[3, 3],
stride=1,
activation_fn=tf.nn.relu,
weights_initializer=cfgs.SUBNETS_WEIGHTS_INITIALIZER,
biases_initializer=cfgs.SUBNETS_BIAS_INITIALIZER,
scope='{}_{}'.format(scope_list[0], i),
reuse=reuse_flag)
add_heatmap(rpn_conv2d_3x3, name='cls_head_%s' % level)
rpn_box_scores = slim.conv2d(rpn_conv2d_3x3,
num_outputs=cfgs.CLASS_NUM * self.num_anchors_per_location,
kernel_size=[3, 3],
stride=1,
weights_initializer=cfgs.SUBNETS_WEIGHTS_INITIALIZER,
biases_initializer=cfgs.FINAL_CONV_BIAS_INITIALIZER,
scope=scope_list[2],
activation_fn=None,
reuse=reuse_flag)
rpn_box_scores = tf.reshape(rpn_box_scores, [-1, cfgs.CLASS_NUM],
name='rpn_{}_classification_reshape'.format(level))
rpn_box_probs = tf.sigmoid(rpn_box_scores, name='rpn_{}_classification_sigmoid'.format(level))
return rpn_box_scores, rpn_box_probs
def rpn_reg_net(self, inputs, scope_list, reuse_flag, level):
rpn_conv2d_3x3 = inputs
for i in range(4):
rpn_conv2d_3x3 = slim.conv2d(
inputs=rpn_conv2d_3x3,
num_outputs=256,
kernel_size=[3, 3],
weights_initializer=cfgs.SUBNETS_WEIGHTS_INITIALIZER,
biases_initializer=cfgs.SUBNETS_BIAS_INITIALIZER,
stride=1,
activation_fn=tf.nn.relu,
scope='{}_{}'.format(scope_list[1], i),
reuse=reuse_flag)
add_heatmap(rpn_conv2d_3x3, name='reg_head_%s' % level)
rpn_delta_boxes = slim.conv2d(
rpn_conv2d_3x3,
num_outputs=5 * self.num_anchors_per_location,
kernel_size=[3, 3],
stride=1,
weights_initializer=cfgs.SUBNETS_WEIGHTS_INITIALIZER,
biases_initializer=cfgs.SUBNETS_BIAS_INITIALIZER,
scope=scope_list[3],
activation_fn=None,
reuse=reuse_flag)
rpn_head_cls = slim.conv2d(
rpn_conv2d_3x3,
num_outputs=4 * self.num_anchors_per_location,
kernel_size=[3, 3],
stride=1,
weights_initializer=cfgs.SUBNETS_WEIGHTS_INITIALIZER,
biases_initializer=cfgs.SUBNETS_BIAS_INITIALIZER,
scope=scope_list[4],
activation_fn=None,
reuse=reuse_flag)
rpn_angle_cls = slim.conv2d(
rpn_conv2d_3x3,
num_outputs=self.angle_range * self.num_anchors_per_location,
kernel_size=[3, 3],
stride=1,
weights_initializer=cfgs.SUBNETS_WEIGHTS_INITIALIZER,
biases_initializer=cfgs.SUBNETS_BIAS_INITIALIZER,
scope=scope_list[5],
activation_fn=None,
reuse=reuse_flag)
rpn_delta_boxes = tf.reshape(
rpn_delta_boxes, [-1, 5],
name='rpn_{}_regression_reshape'.format(level))
rpn_head_cls = tf.reshape(rpn_head_cls, [-1, 4],
name='rpn_{}_head_cls_reshape'.format(level))
rpn_angle_cls = tf.reshape(
rpn_angle_cls, [-1, self.angle_range],
name='rpn_{}_angle_cls_reshape'.format(level))
return rpn_delta_boxes, rpn_head_cls, rpn_angle_cls
def resnet_base(img_batch, scope_name, is_training=True):
if scope_name.endswith('b'):
get_resnet_fn = get_resnet_v1_b_base
elif scope_name.endswith('d'):
get_resnet_fn = get_resnet_v1_d_base
else:
raise ValueError("scope Name erro....")
_, feature_dict = get_resnet_fn(
input_x=img_batch,
scope=scope_name,
bottleneck_nums=BottleNeck_NUM_DICT[scope_name],
base_channels=BASE_CHANNELS_DICT[scope_name],
is_training=is_training,
freeze_norm=True,
freeze=cfgs.FREEZE_BLOCKS)
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, 2, -1):
add_heatmap(pyramid_dict['P%d' % level],
name='Layer%d/P%d_heat' % (level, level))
return pyramid_dict