FPN在faster_rcnn中实现细节代码说明

小程序：扫一扫查出行
【扫一扫了解最新限行尾号】
复制小程序

代码参考自：https://github.com/DetectionTeamUCAS/FPN_Tensorflow

主要分析fpn多层金字塔结构的输出如何进行预测。

FPN金字塔结构插入在faster_rcnn的特征图获取之后，在rpn结构之前。

具体代码如下所示：

代码结构追溯至FPN部分：

train.py(line 46 :build_whole_detection_network函数）

build_whole_network(line 372: build_whole_detection_network函数）

按照注释分别查看7个步骤：

1. build base network

2. build rpn

3. generate_anchors

4. postprocess rpn proposals. such as: decode, clip, NMS(所得第一次框处理)

5. build Fast-RCNN（5，roipooling 6,inference rois to obtain fc 7,cls reg）

6. postprocess_fastrcnn(最后框处理)
FPN部分在build base network中，得到的plist即为金字塔特征图的集合
build_base_network一步一步回溯找到原函数resnet_base(fpn操作在这里，如下代码)

def resnet_base(img_batch, scope_name, is_training=True):
'''
this code is derived from light-head rcnn.
https://github.com/zengarden/light_head_rcnn It is convenient to freeze blocks. So we adapt this mode.
'''
if scope_name == 'resnet_v1_50':
middle_num_units = 6
elif scope_name == 'resnet_v1_101':
middle_num_units = 23
else:
raise NotImplementedError('We only support resnet_v1_50 or resnet_v1_101. Check your network name....yjr') blocks = [resnet_v1_block('block1', base_depth=64, num_units=3, stride=2),
resnet_v1_block('block2', base_depth=128, num_units=4, stride=2),
resnet_v1_block('block3', base_depth=256, num_units=middle_num_units, stride=2),
resnet_v1_block('block4', base_depth=512, num_units=3, stride=1)]
# when use fpn . stride list is [1, 2, 2] with slim.arg_scope(resnet_arg_scope(is_training=False)):
with tf.variable_scope(scope_name, scope_name):
# Do the first few layers manually, because 'SAME' padding can behave inconsistently
# for images of different sizes: sometimes 0, sometimes 1
net = resnet_utils.conv2d_same(
img_batch, 64, 7, stride=2, scope='conv1')
net = tf.pad(net, [[0, 0], [1, 1], [1, 1], [0, 0]])
net = slim.max_pool2d(
net, [3, 3], stride=2, padding='VALID', scope='pool1') not_freezed = [False] * cfgs.FIXED_BLOCKS + (4-cfgs.FIXED_BLOCKS)*[True]
# Fixed_Blocks can be 1~3 with slim.arg_scope(resnet_arg_scope(is_training=(is_training and not_freezed[0]))):
C2, end_points_C2 = resnet_v1.resnet_v1(net,
blocks[0:1],
global_pool=False,
include_root_block=False,
scope=scope_name) # C2 = tf.Print(C2, [tf.shape(C2)], summarize=10, message='C2_shape')
add_heatmap(C2, name='Layer2/C2_heat') with slim.arg_scope(resnet_arg_scope(is_training=(is_training and not_freezed[1]))):
C3, end_points_C3 = resnet_v1.resnet_v1(C2,
blocks[1:2],
global_pool=False,
include_root_block=False,
scope=scope_name) # C3 = tf.Print(C3, [tf.shape(C3)], summarize=10, message='C3_shape')
add_heatmap(C3, name='Layer3/C3_heat')
with slim.arg_scope(resnet_arg_scope(is_training=(is_training and not_freezed[2]))):
C4, end_points_C4 = resnet_v1.resnet_v1(C3,
blocks[2:3],
global_pool=False,
include_root_block=False,
scope=scope_name) add_heatmap(C4, name='Layer4/C4_heat') # C4 = tf.Print(C4, [tf.shape(C4)], summarize=10, message='C4_shape')
with slim.arg_scope(resnet_arg_scope(is_training=is_training)):
C5, end_points_C5 = resnet_v1.resnet_v1(C4,
blocks[3:4],
global_pool=False,
include_root_block=False,
scope=scope_name)
# C5 = tf.Print(C5, [tf.shape(C5)], summarize=10, message='C5_shape')
add_heatmap(C5, name='Layer5/C5_heat') feature_dict = {'C2': end_points_C2['{}/block1/unit_2/bottleneck_v1'.format(scope_name)],
'C3': end_points_C3['{}/block2/unit_3/bottleneck_v1'.format(scope_name)],
'C4': end_points_C4['{}/block3/unit_{}/bottleneck_v1'.format(scope_name, middle_num_units - 1)],
'C5': end_points_C5['{}/block4/unit_3/bottleneck_v1'.format(scope_name)],
# 'C5': end_points_C5['{}/block4'.format(scope_name)],
} # feature_dict = {'C2': C2,
# 'C3': C3,
# 'C4': C4,
# 'C5': C5} 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(C5,
num_outputs=256,
kernel_size=[1, 1],
stride=1, scope='build_P5')
if "P6" in cfgs.LEVLES:
P6 = slim.max_pool2d(P5, kernel_size=[1, 1], stride=2, scope='build_P6')
pyramid_dict['P6'] = P6 pyramid_dict['P5'] = P5 for level in range(4, 1, -1): # build [P4, P3, P2] 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(4, 1, -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)
for level in range(5, 1, -1):
add_heatmap(pyramid_dict['P%d' % level], name='Layer%d/P%d_heat' % (level, level)) # return [P2, P3, P4, P5, P6]
print("we are in Pyramid::-======>>>>")
print(cfgs.LEVLES)
print("base_anchor_size are: ", cfgs.BASE_ANCHOR_SIZE_LIST)
print(20 * "__")
return [pyramid_dict[level_name] for level_name in cfgs.LEVLES]
# return pyramid_dict # return the dict. And get each level by key. But ensure the levels are consitant
# return list rather than dict, to avoid dict is unordered

观察原特征图的结构C2,C3,C4,C5, 以及特征金字塔的结构P5,P4,P3,P2，为5层的特征金字塔结构。

操作如图：

金字塔结构的总层数为（p5,p6,p4,p3,p2)

P5 = conv2d(C5) (因金字塔特征图每层的构造是上面一层的2x upsaming 和左边的1*1conv后的结果相加)

P6 = max_pool(P5)

核心的融合部分在下面代码中显示：

P4 = C4 + P5

P3 = C3 + P4

P2 = C2 + P3

for level in range(4, 1, -1): # build [P4, P3, P2] 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)

最后的P_LIST共有：LEVLES = ['P2', 'P3', 'P4', 'P5', 'P6']层级

对应每层特征图设置不同大小的anchors

Instead, we assign anchors of a single scale to each level. Formally, we deﬁne the anchors to have areas of {32,64,128,256,512} pixels on {P2,P3,P4,P5,P6} respectively.

As in [29] we also use anchors of multiple aspect ratios{1:2, 1:1, 2:1}at each level. So in total there are anchors over the pyramid

后面得到全部金字塔特征图的roi，下一步是要把roi对应到各自层的特征图上取roi特征，不同大小的roi对应不同的特征图，较大的roi对应深层的特征图，按照公式