def darknet53(input_data):
input_data = common.convolutional(input_data, (3, 3, 3, 32))
input_data = common.convolutional(input_data, (3, 3, 32, 64),
downsample=True)
for i in range(1):
input_data = common.residual_block(input_data, 64, 32, 64)
input_data = common.convolutional(input_data, (3, 3, 64, 128),
downsample=True)
for i in range(2):
input_data = common.residual_block(input_data, 128, 64, 128)
input_data = common.convolutional(input_data, (3, 3, 128, 256),
downsample=True)
for i in range(8):
input_data = common.residual_block(input_data, 256, 128, 256)
route_1 = input_data
input_data = common.convolutional(input_data, (3, 3, 256, 512),
downsample=True)
for i in range(8):
input_data = common.residual_block(input_data, 512, 256, 512)
route_2 = input_data
input_data = common.convolutional(input_data, (3, 3, 512, 1024),
downsample=True)
for i in range(4):
input_data = common.residual_block(input_data, 1024, 512, 1024)
return route_1, route_2, input_data
def resnet(input_data, num_layer):
if num_layer == 18:
restnet = ResNet18()
if num_layer == 34:
restnet = ResNet34()
if num_layer == 50:
restnet = ResNet50()
if num_layer == 101:
restnet = ResNet101()
if num_layer == 152:
restnet = ResNet152()
output_resnet = restnet.call(input_data, training=True)
route_1 = common.convolutional(output_resnet, (3, 3, 2048, 256), downsample=False)
for _ in range(8):
route_1 = common.residual_block(route_1, 256, 128, 256)
route_2 = common.convolutional(output_resnet, (3, 3, 2048, 512), downsample=True)
for _ in range(8):
route_2 = common.residual_block(route_2, 512, 256, 512)
route_3 = common.convolutional(route_2, (3, 3, 512, 1024), downsample=True)
for _ in range(4):
route_3 = common.residual_block(route_3, 1024, 512, 1024)
return route_1, route_2, route_3
def darknet53(input_data):
'''
input_data是3通道的
output_data是1024通道的
'''
x = common.convolutional(input_data, (3, 3, 3, 32))
x = common.convolutional(x, (3, 3, 32, 64), downsample='True')
for i in range(1):
x = common.residual_block(x, 64, 32, 64)
x = common.convolutional(x, (3, 3, 64, 128), downsample='True')
for i in range(2):
x = common.residual_block(x, 128, 64, 128)
x = common.convolutional(x, (3, 3, 128, 256), downsample='True')
for i in range(8):
x = common.residual_block(x, 256, 128, 256)
route_1 = x
x = common.convolutional(x, (3, 3, 256, 512), downsample='True')
for i in range(8):
x = common.residual_block(x, 512, 256, 512)
route_2 = x
output_data = common.convolutional(x, (3, 3, 512, 1024), downsample='True')
return route_1, route_2, output_data
def darknet53(input_data, trainable):
with tf.variable_scope('darknet'):
#filters_shape=(3, 3, 3, 32),前2个3代表卷积核大小，第三个3代表输入图像的通道数数，此处为RGB，32是输出的通道数
input_data = common.convolutional(input_data, filters_shape=(3, 3, 3, 32), trainable=trainable, name='conv0')
#conv0 416*416*3 ---- 416*416*32
input_data = common.convolutional(input_data, filters_shape=(3, 3, 32, 64),
trainable=trainable, name='conv1', downsample=True)
# conv1 416*416*32 ---- 208*208*64
for i in range(1):
input_data = common.residual_block(input_data, 64, 32, 64, trainable=trainable, name='residual%d' %(i+0))
#residual 0 208*208*64---208*208*64
input_data = common.convolutional(input_data, filters_shape=(3, 3, 64, 128),
trainable=trainable, name='conv4', downsample=True)
# conv4 208*208*64---104*104*128
for i in range(2):
input_data = common.residual_block(input_data, 128, 64, 128, trainable=trainable, name='residual%d' %(i+1))
# residual 1 104*104*128---104*104*128
# residual 2 104*104*128---104*104*128
route_1 = input_data #4倍下采样的输出，104*104*128
input_data = common.convolutional(input_data, filters_shape=(3, 3, 128, 256),
trainable=trainable, name='conv9', downsample=True)
# conv9 104*104*128---52*52*256
for i in range(8):
input_data = common.residual_block(input_data, 256, 128, 256, trainable=trainable, name='residual%d' %(i+3))
# residual 3 52*52*256---52*52*128
# residual 4 52*52*128---52*52*256
# residual 5 52*52*256---52*52*128
# residual 6 52*52*128---52*52*256
# residual 7 52*52*256---52*52*128
# residual 8 52*52*128---52*52*256
# residual 9 52*52*256---52*52*128
# residual 10 52*52*128---52*52*256
route_2 = input_data #用作后面的concat操作，是8倍下采样的输出52*52*256
input_data = common.convolutional(input_data, filters_shape=(3, 3, 256, 512),
trainable=trainable, name='conv26', downsample=True)
# conv26 52*52*256---26*26*512
for i in range(8):
input_data = common.residual_block(input_data, 512, 256, 512, trainable=trainable, name='residual%d' %(i+11))
# residual 11 26*26*512---26*26*256
# residual 12 26*26*256---26*26*512
# residual 13 26*26*512---26*26*256
# residual 14 26*26*256---26*26*512
# residual 15 26*26*512---26*26*256
# residual 16 26*26*256---26*26*512
# residual 17 26*26*512---26*26*256
# residual 18 26*26*256---26*26*512
route_3 = input_data #16倍下采样的输出26*26*512
input_data = common.convolutional(input_data, filters_shape=(3, 3, 512, 1024),
trainable=trainable, name='conv43', downsample=True)
# conv43 26*26*512---13*13*1024
for i in range(4):
input_data = common.residual_block(input_data, 1024, 512, 1024, trainable=trainable, name='residual%d' %(i+19))
# residual 19 13*13*1024---13*13*512
# residual 20 13*13*512---13*13*1024
return route_1, route_2, route_3, input_data
def darknet53(input_data, trainable):
with tf.variable_scope('darknet'):
output_data = common.convolutional(input_data, filter_shape=(3,3,32,32),
trainable=trainable, name='conv1')
output_data = common.convolutional(output_data, filter_shape=(3,3,32,64),
trainable=trainable, downsample=True, name='conv2')
for i in range(1):
output_data = common.residual_block(output_data, 64, 32, 64,
trainable=trainable, name='residual%d'%(i+1))
output_data = common.convolutional(output_data, filter_shape=(3,3,64,128),
trainable=trainable, downsample=True, name='conv5')
for i in range(2):
output_data = common.residual_block(output_data, 128, 64, 128,
trainable=trainable, name='residual%d'%(i+2))
output_data = common.convolutional(output_data, filter_shape=(3,3,128,256),
trainable=trainable, downsample=True, name='conv10')
for i in range(8):
output_data = common.residual_block(output_data, 256, 128, 256,
trainable=trainable, name='residual%d'%(i+4))
route_1 = output_data
output_data = common.convolutional(output_data, filter_shape=(3,3,256,512),
trainable=trainable, downsample=True, name='conv27')
for i in range(8):
output_data = common.residual_block(output_data, 512, 256, 512,
trainable=trainable, name='residual%d'%(i+12))
route_2=output_data
output_data = common.convolutional(output_data, filter_shape=(3,3,512,1024),
trainable=trainable, downsample=True, name='conv44')
for i in range(4):
output_data = common.residual_block(output_data, 1024, 512, 1024,
trainable=trainable, name='residual%d'%(i+20))
return route_1, route_2, output_data
def darknet53(input_data, trainable):
with tf.variable_scope('darknet'):
input_data = common.convolutional(input_data, filters_shape=(3, 3, 3, 32), trainable=trainable, name='conv0')
input_data = common.convolutional(input_data, filters_shape=(3, 3, 32, 64),
trainable=trainable, name='conv1', downsample=True)
for i in range(1):
input_data = common.residual_block(input_data, 64, 32, 64, trainable=trainable, name='residual%d' % (i + 0))
input_data = common.convolutional(input_data, filters_shape=(3, 3, 64, 128),
trainable=trainable, name='conv4', downsample=True)
for i in range(2):
input_data = common.residual_block(input_data, 128, 64, 128, trainable=trainable,
name='residual%d' % (i + 1))
input_data = common.convolutional(input_data, filters_shape=(3, 3, 128, 256),
trainable=trainable, name='conv9', downsample=True)
for i in range(8):
input_data = common.residual_block(input_data, 256, 128, 256, trainable=trainable,
name='residual%d' % (i + 3))
route_1 = input_data
input_data = common.convolutional(input_data, filters_shape=(3, 3, 256, 512),
trainable=trainable, name='conv26', downsample=True)
for i in range(8):
input_data = common.residual_block(input_data, 512, 256, 512, trainable=trainable,
name='residual%d' % (i + 11))
route_2 = input_data
input_data = common.convolutional(input_data, filters_shape=(3, 3, 512, 1024),
trainable=trainable, name='conv43', downsample=True)
for i in range(4):
input_data = common.residual_block(input_data, 1024, 512, 1024, trainable=trainable,
name='residual%d' % (i + 19))
return route_1, route_2, input_data
def darknet53(input_data):
"""
build backbone according to the paper
"""
input_data = common.convolutional(input_data, filters_shape=(3, 3, 3, 32))
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 32, 64),
downsample=True)
for _ in range(1):
input_data = common.residual_block(input_data, 32, 64)
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 64, 128),
downsample=True)
for _ in range(2):
input_data = common.residual_block(input_data, 64, 128)
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 128, 256),
downsample=True)
for _ in range(8):
input_data = common.residual_block(input_data, 128, 256)
route1 = input_data
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 256, 512),
downsample=True)
for _ in range(8):
input_data = common.residual_block(input_data, 256, 512)
route2 = input_data
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 512, 1024),
downsample=True)
for _ in range(4):
input_data = common.residual_block(input_data, 512, 1024)
return route1, route2, input_data
def darknet53(input_data):
'''
If downsample=True, the feature map size will be decreased by a factor of 2.
'''
input_data = common.convolutional(input_data, (3, 3, 3, 32))
input_data = common.convolutional(input_data, (3, 3, 32, 64),
downsample=True)
for i in range(1):
input_data = common.residual_block(input_data, 64, 32, 64)
input_data = common.convolutional(input_data, (3, 3, 64, 128),
downsample=True)
for i in range(2):
input_data = common.residual_block(input_data, 128, 64, 128)
input_data = common.convolutional(input_data, (3, 3, 128, 256),
downsample=True)
for i in range(8):
input_data = common.residual_block(input_data, 256, 128, 256)
route_1 = input_data
input_data = common.convolutional(input_data, (3, 3, 256, 512),
downsample=True)
for i in range(8):
input_data = common.residual_block(input_data, 512, 256, 512)
route_2 = input_data
input_data = common.convolutional(input_data, (3, 3, 512, 1024),
downsample=True)
for i in range(4):
input_data = common.residual_block(input_data, 1024, 512, 1024)
return route_1, route_2, input_data
def darknet53(input_data):
#conv2d Filters=32 Size=3x3 Strike=1
input_data = common.convolutional(input_data, (3, 3, 3, 32)) # 256
#conv2d Filters=64 Size=3x3 Strike=2
input_data = common.convolutional(input_data, (3, 3, 32, 64), downsample=True) # 128
# The first block residual
for _ in range(1):
input_data = common.residual_block(input_data, 64, 32, 64)
input_data = common.convolutional(input_data, (3, 3, 64, 128), downsample=True) # 64
for _ in range(2):
input_data = common.residual_block(input_data, 128, 64, 128)
input_data = common.convolutional(input_data, (3, 3, 128, 256), downsample=True) # 32
for _ in range(8):
input_data = common.residual_block(input_data, 256, 128, 256)
route_1 = input_data # 256
input_data = common.convolutional(input_data, (3, 3, 256, 512), downsample=True) # 16
for _ in range(8):
input_data = common.residual_block(input_data, 512, 256, 512)
route_2 = input_data #512
input_data = common.convolutional(input_data, (3, 3, 512, 1024), downsample=True) # 8
for _ in range(4):
input_data = common.residual_block(input_data, 1024, 512, 1024)
print('route_1: ', route_1.shape)
print('route_2: ', route_2.shape)
print('input_data: ', input_data.shape)
return route_1, route_2, input_data # 52 | 26 | 13
def darknet53(input_rgb,input_lwir, trainable):
with tf.variable_scope('darknet'):
##############################RGB###############################################
input_rgb = common.convolutional(input_rgb, filters_shape=(3, 3, 3, 32), trainable=trainable, name='conv0')
input_rgb = common.convolutional(input_rgb, filters_shape=(3, 3, 32, 64),
trainable=trainable, name='conv1', downsample=True)
for i in range(1):
input_rgb = common.residual_block(input_rgb, 64, 32, 64, trainable=trainable, name='residual%d' %(i+0))
input_rgb = common.convolutional(input_rgb, filters_shape=(3, 3, 64, 128),
trainable=trainable, name='conv4', downsample=True)
for i in range(2):
input_rgb = common.residual_block(input_rgb, 128, 64, 128, trainable=trainable, name='residual%d' %(i+1))
input_rgb = common.convolutional(input_rgb, filters_shape=(3, 3, 128, 256),
trainable=trainable, name='conv9', downsample=True)
##############################RGB###############################################
##############################LWIR###############################################
input_lwir = common.convolutional(input_lwir, filters_shape=(3, 3, 3, 32), trainable=trainable, name='conv0_lwir')
input_lwir = common.convolutional(input_lwir, filters_shape=(3, 3, 32, 64),
trainable=trainable, name='conv1_lwir', downsample=True)
for i in range(1):
input_lwir = common.residual_block(input_lwir, 64, 32, 64, trainable=trainable, name='residual%d_lwir' %(i+0))
input_lwir = common.convolutional(input_lwir, filters_shape=(3, 3, 64, 128),
trainable=trainable, name='conv4_lwir', downsample=True)
for i in range(2):
input_lwir = common.residual_block(input_lwir, 128, 64, 128, trainable=trainable, name='residual%d_lwir' %(i+1))
input_lwir = common.convolutional(input_lwir, filters_shape=(3, 3, 128, 256),
trainable=trainable, name='conv9_lwir', downsample=True)
##############################LWIR###############################################
##############################cont###############################################
input_data = tf.concat(axis=3, values=[input_rgb, input_lwir])
input_data = common.convolutional(input_data, filters_shape=(3, 3, 512, 256),trainable=trainable, name='conv_b09')
##############################cont###############################################
for i in range(8):
input_data = common.residual_block(input_data, 256, 128, 256, trainable=trainable, name='residual%d' %(i+3))
route_1 = input_data
input_data = common.convolutional(input_data, filters_shape=(3, 3, 256, 512),
trainable=trainable, name='conv26', downsample=True)
for i in range(8):
input_data = common.residual_block(input_data, 512, 256, 512, trainable=trainable, name='residual%d' %(i+11))
route_2 = input_data
input_data = common.convolutional(input_data, filters_shape=(3, 3, 512, 1024),
trainable=trainable, name='conv43', downsample=True)
for i in range(4):
input_data = common.residual_block(input_data, 1024, 512, 1024, trainable=trainable, name='residual%d' %(i+19))
return route_1, route_2, input_data
def mobilenetv2(input_data, trainable):
with tf.variable_scope('mobilenetv2'):
#input_data = tf.reshape(input_data, [-1, 608, 608, 3]) # print layer's shape
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 3, 32),
trainable=trainable,
name='conv0',
downsample=True)
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 32, 16),
trainable=trainable,
name='conv1',
downsample=True)
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 16, 24),
trainable=trainable,
name='conv2')
for i in range(1):
input_data = common.residual_block(input_data,
24,
24,
24,
trainable=trainable,
name='residual%d' % (i + 0))
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 24, 32),
trainable=trainable,
name='conv4',
downsample=True)
for i in range(2):
input_data = common.residual_block(input_data,
32,
32,
32,
trainable=trainable,
name='residual%d' % (i + 1))
route_1 = input_data
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 32, 64),
trainable=trainable,
name='conv7',
downsample=True)
for i in range(3):
input_data = common.residual_block(input_data,
64,
384,
64,
trainable=trainable,
name='residual%d' % (i + 3))
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 64, 96),
trainable=trainable,
name='conv11')
for i in range(2):
input_data = common.residual_block(input_data,
96,
576,
96,
trainable=trainable,
name='residual%d' % (i + 6))
route_2 = input_data
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 96, 160),
trainable=trainable,
name='conv14',
downsample=True)
for i in range(2):
input_data = common.residual_block(input_data,
160,
160,
160,
trainable=trainable,
name='residual%d' % (i + 8))
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 160, 320),
trainable=trainable,
name='conv17')
return route_1, route_2, input_data
def darknet53(input_data): # 用了五次下采样，所以图片的大小应该维持32的倍数
with tf.variable_scope('darknet53'):
# 0
input_data = common.convolutional_Layer(input_data,
filters_shape=(3, 3, 3, 32),
name='conv0')
# -------------
# 1
input_data = common.convolutional_Layer(input_data,
filters_shape=(3, 3, 32, 64),
name='conv1',
downsample=True)
# 2-3(1*2)
for i in range(1):
input_data = common.residual_block(input_data,
64,
32,
64,
name='residual%d' % (i + 0))
# -------------
# 4
input_data = common.convolutional_Layer(input_data,
filters_shape=(3, 3, 64, 128),
downsample=True,
name='conv4')
# 5-8(2*2)
for i in range(2):
input_data = common.residual_block(input_data,
128,
64,
128,
name='residual%d' % (i + 1))
# -------------
# 9
input_data = common.convolutional_Layer(input_data,
filters_shape=(3, 3, 128, 256),
downsample=True,
name='conv9')
# 10-25(8*2)
for i in range(8):
input_data = common.residual_block(input_data,
256,
128,
256,
name='residual%d' % (i + 3))
middle_data_1 = input_data
# -------------
# 26
input_data = common.convolutional_Layer(input_data,
filters_shape=(3, 3, 256, 512),
downsample=True,
name='conv26')
# 27-42(8*2)
for i in range(8):
input_data = common.residual_block(input_data,
512,
256,
512,
name='residual%d' % (i + 11))
middle_data_2 = input_data
# -------------
# 43
input_data = common.convolutional_Layer(input_data,
filters_shape=(3, 3, 512,
1024),
downsample=True,
name='conv43')
# 44-51(4*2)
for i in range(4):
input_data = common.residual_block(input_data,
1024,
512,
1024,
name='residual%d' % (i + 19))
# 备注 darknet中还有三层（Avgpool、Connected 和 softmax layer）
# 用于在 Imagenet 数据集上作分类训练用的，yolo是另一种思路，是不会用到这三层的，只是用前面来特征提取，谁让darknet53这么给力捏。
return middle_data_1, middle_data_2, input_data
def darknet53(input_data, trainable):
"""
基于输入数据构建Darknet53的网络
:param input_data:
:param trainable:
:return:
"""
with tf.variable_scope('darknet'):
# 1. 第一层卷积[3,3,3,32]/1
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 3, 32),
trainable=trainable,
name='conv0')
# 2. 第二层卷积[3,3,32,64]/2
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 32, 64),
trainable=trainable,
name='conv1',
downsample=True)
# 3. 进入残差+卷积的结构
# a. 第一个残差结构
for i in range(1):
# 残差结构，通道数变化情况: 64 -> 32 -> 64
input_data = common.residual_block(input_data,
64,
32,
64,
trainable=trainable,
name='residual%d' % (i + 0))
# b. 进入残差后的卷积结构(下采样), [3,3,64,128]/2
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 64, 128),
trainable=trainable,
name='conv4',
downsample=True)
# c. 进入第二组残差结构
for i in range(2):
# 残差结构, 通道数变化情况: 128->64->128
input_data = common.residual_block(input_data,
128,
64,
128,
trainable=trainable,
name='residual%d' % (i + 1))
# d. 进入残差后的卷积结构(下采样), [3,3,128,256]/2
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 128, 256),
trainable=trainable,
name='conv9',
downsample=True)
# e. 进入第三组残差结构
for i in range(8):
# 残差结构, 通道数变化情况: 256->128->256
input_data = common.residual_block(input_data,
256,
128,
256,
trainable=trainable,
name='residual%d' % (i + 3))
# f. 定义第一个输出分支
route_1 = input_data
# g. 进入残差后的卷积结构(下采样), [3,3,256,512]/2
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 256, 512),
trainable=trainable,
name='conv26',
downsample=True)
# h. 进入第四组残差结构
for i in range(8):
# 残差结构, 通道数变化情况: 512->256->512
input_data = common.residual_block(input_data,
512,
256,
512,
trainable=trainable,
name='residual%d' % (i + 11))
# i. 定义第二个分支输出
route_2 = input_data
# j. 进入残差后的卷积结构(下采样), [3,3,512,1024]/2
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 512, 1024),
trainable=trainable,
name='conv43',
downsample=True)
# k. 进入第五组残差结构
for i in range(4):
# 残差结构, 通道数变化情况: 1024->512->1024
input_data = common.residual_block(input_data,
1024,
512,
1024,
trainable=trainable,
name='residual%d' % (i + 19))
# 返回第一分支、第二分支以及最终的结构数据
return route_1, route_2, input_data
def cspdarknet53(input_data):
print(
'[INFO][core.backbone.cspdarknet53] Initializing cspdarknet53 backbone-architecture '
)
input_data = common.convolutional(input_data, (3, 3, 3, 32),
activate_type="mish")
input_data = common.convolutional(input_data, (3, 3, 32, 64),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 64, 64), activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 64, 64),
activate_type="mish")
for i in range(1):
input_data = common.residual_block(input_data,
64,
32,
64,
activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 64, 64),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 128, 64),
activate_type="mish")
input_data = common.convolutional(input_data, (3, 3, 64, 128),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 128, 64), activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 128, 64),
activate_type="mish")
for i in range(2):
input_data = common.residual_block(input_data,
64,
64,
64,
activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 64, 64),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 128, 128),
activate_type="mish")
input_data = common.convolutional(input_data, (3, 3, 128, 256),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 256, 128), activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 256, 128),
activate_type="mish")
for i in range(8):
input_data = common.residual_block(input_data,
128,
128,
128,
activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 128, 128),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 256, 256),
activate_type="mish")
route_1 = input_data
input_data = common.convolutional(input_data, (3, 3, 256, 512),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 512, 256), activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 512, 256),
activate_type="mish")
for i in range(8):
input_data = common.residual_block(input_data,
256,
256,
256,
activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 256, 256),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 512, 512),
activate_type="mish")
route_2 = input_data
input_data = common.convolutional(input_data, (3, 3, 512, 1024),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 1024, 512),
activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 1024, 512),
activate_type="mish")
for i in range(4):
input_data = common.residual_block(input_data,
512,
512,
512,
activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 512, 512),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 1024, 1024),
activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 1024, 512))
input_data = common.convolutional(input_data, (3, 3, 512, 1024))
input_data = common.convolutional(input_data, (1, 1, 1024, 512))
input_data = tf.concat([
tf.nn.max_pool(input_data, ksize=13, padding='SAME', strides=1),
tf.nn.max_pool(input_data, ksize=9, padding='SAME', strides=1),
tf.nn.max_pool(input_data, ksize=5, padding='SAME', strides=1),
input_data
],
axis=-1)
input_data = common.convolutional(input_data, (1, 1, 2048, 512))
input_data = common.convolutional(input_data, (3, 3, 512, 1024))
input_data = common.convolutional(input_data, (1, 1, 1024, 512))
print(
'[INFO][core.backbone.cspdarknet53] cspdarknet53 backbone-architecture initialized'
)
return route_1, route_2, input_data
def darknet53(input_data, trainable):
with tf.variable_scope('darknet'):
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 3, 32),
trainable=trainable,
name='conv0')
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 32, 64),
trainable=trainable,
name='conv1',
downsample=True)
# 依次为五个残差块 1-2-8-8-4
for i in range(1):
# 64 32 64 依次为上一输出的通道数 残差中两步卷积的通道
input_data = common.residual_block(input_data,
64,
32,
64,
trainable=trainable,
name='residual%d' % (i + 0))
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 64, 128),
trainable=trainable,
name='conv4',
downsample=True)
for i in range(2):
input_data = common.residual_block(input_data,
128,
64,
128,
trainable=trainable,
name='residual%d' % (i + 1))
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 128, 256),
trainable=trainable,
name='conv9',
downsample=True)
for i in range(8):
input_data = common.residual_block(input_data,
256,
128,
256,
trainable=trainable,
name='residual%d' % (i + 3))
route_1 = input_data
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 256, 512),
trainable=trainable,
name='conv26',
downsample=True)
for i in range(8):
input_data = common.residual_block(input_data,
512,
256,
512,
trainable=trainable,
name='residual%d' % (i + 11))
route_2 = input_data
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 512, 1024),
trainable=trainable,
name='conv43',
downsample=True)
for i in range(4):
input_data = common.residual_block(input_data,
1024,
512,
1024,
trainable=trainable,
name='residual%d' % (i + 19))
# route1 2 3 inpudata 分别为用于后面concat的值 即 128残差后用于最大特征图concat的 route1 以此类推
return route_1, route_2, input_data
def darknet53(input_data, trainable):
with tf.variable_scope('darknet'):
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 3, 32),
trainable=trainable,
name='conv0') # 416 x 416 x 32
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 32, 64),
trainable=trainable,
name='conv1',
downsample=True) # 208 x 208 x 64
for i in range(1):
residual_data = common.residual_block(
input_data,
64,
32,
64,
trainable=trainable,
name='residual%d' %
(i + 0)) # 208 x 208 x 32, 208 x 208 x 64
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
input_data = common.attention_residual_block(
input_data,
residual_data,
208,
64,
name='attention_residual%d' % (i + 0))
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 64, 128),
trainable=trainable,
name='conv4',
downsample=True) # 104 x 104 x 128
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
for i in range(2):
residual_data = common.residual_block(
input_data,
128,
64,
128,
trainable=trainable,
name='residual%d' %
(i + 1)) # 104 x 104 x 64, 104 x 104 x 128
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
input_data = common.attention_residual_block(
input_data,
residual_data,
104,
128,
name='attention_residual%d' % (i + 1))
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 128, 256),
trainable=trainable,
name='conv9',
downsample=True) # 52 x 52 x 256
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
for i in range(8):
residual_data = common.residual_block(
input_data,
256,
128,
256,
trainable=trainable,
name='residual%d' % (i + 3)) # 52 x 52 x 128, 52 x 52 x 256
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
input_data = common.attention_residual_block(
input_data,
residual_data,
52,
256,
name='attention_residual%d' % (i + 3))
route_1 = input_data
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 256, 512),
trainable=trainable,
name='conv26',
downsample=True) # 26 x 26 x 512
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
for i in range(8):
residual_data = common.residual_block(
input_data,
512,
256,
512,
trainable=trainable,
name='residual%d' % (i + 11)) # 26 x 26 x 256, 26 x 26 x 512
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
input_data = common.attention_residual_block(
input_data,
residual_data,
26,
512,
name='attention_residual%d' % (i + 11))
route_2 = input_data
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 512, 1024),
trainable=trainable,
name='conv43',
downsample=True) # 13 x 13 x1024
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
for i in range(4):
residual_data = common.residual_block(
input_data,
1024,
512,
1024,
trainable=trainable,
name='residual%d' % (i + 19)) # 13 x 13 x 512, 13 x 13 x1024
input_data = tf.layers.dropout(input_data,
rate=0.5,
training=trainable)
input_data = common.attention_residual_block(
input_data,
residual_data,
13,
1024,
name='attention_residual%d' % (i + 19))
return route_1, route_2, input_data
def darknet53(input_data, trainable):
'''
53层卷积网络
:param input_dat:
:param trainable:
:return:
'''
with tf.variable_scope('darknet'):
input_data = common.convolutional(
input_data,
filters_shape=(3, 3, 3, 32), #3X3卷积，3通道，32个卷积核
trainable=trainable,
name='conv0')
input_data = common.convolutional(
input_data,
filters_shape=(3, 3, 32, 64), #输入数据厚度为32，输出厚度为64
trainable=trainable,
name='conv1',
downsample=True) #第1次下采样
for i in range(1):
input_data = common.residual_block(input_data,
64,
32,
64,
trainable=trainable,
name='residual%d' %
(i + 0)) #输出厚度为64
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 64, 128),
trainable=trainable,
name='conv4',
downsample=True) #第2次下采样
for i in range(2):
input_data = common.residual_block(input_data,
128,
64,
128,
trainable=trainable,
name='residual%d' %
(i + 1)) #输出厚度为128
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 128, 256),
trainable=trainable,
name='conv9',
downsample=True) #第3次下采样
for i in range(8):
input_data = common.residual_block(input_data,
256,
128,
256,
trainable=trainable,
name='residual%d' %
(i + 3)) #输出厚度为256
route_1 = input_data #52X52 52来源：输入数据大小为416X416，至此，经过了3次下采样，每下采样一次，大小缩小一半，故3次下采样后，大小为416/(2^3)=52
input_data = common.convolutional(input_data,
filters_shape=(3, 3, 256, 512),
trainable=trainable,
name='conv26',
downsample=True) #第4次下采样
for i in range(8):
input_data = common.residual_block(input_data,
512,
256,
512,
trainable=trainable,
name='residual%d' %
(i + 11)) #输出厚度为512
route_2 = input_data #26X26
input_data = common.convolutional(
input_data,
filters_shape=(3, 3, 512, 1024),
trainable=trainable,
name='conv43',
downsample=True
) #第5次下采样,该次下采样之后，input_data.shape=(?,13,13,1024).?与批次输入的图片数量有关
for i in range(4):
input_data = common.residual_block(input_data,
1024,
512,
1024,
trainable=trainable,
name='residual%d' % (i + 19))
#input_data:13X13
return route_1, route_2, input_data
def darknet53(input_data):
"""
build the darknet 53
:param input_data: [416,416,3]
:return: 3 feature maps [52,52,256], [26,26,512], [13,13,1024]
"""
input_data = common.convolutional(
input_data, (3, 3, 3, 32),
conv_trainable=False) # 32 filters [3,3,3], output [416,416,32]
input_data = common.convolutional(
input_data, (3, 3, 32, 64), downsample=True,
conv_trainable=False) # 64 filters [3,3,3], output [208, 208, 64]
for i in range(1):
input_data = common.residual_block(input_data,
64,
32,
64,
conv_trainable=False)
input_data = common.convolutional(input_data, (3, 3, 64, 128),
downsample=True,
conv_trainable=False)
for i in range(2):
input_data = common.residual_block(input_data,
128,
64,
128,
conv_trainable=False)
input_data = common.convolutional(input_data, (3, 3, 128, 256),
downsample=True,
conv_trainable=False)
for i in range(8):
input_data = common.residual_block(input_data,
256,
128,
256,
conv_trainable=False)
route_1 = input_data
input_data = common.convolutional(input_data, (3, 3, 256, 512),
downsample=True,
conv_trainable=False)
for i in range(8):
input_data = common.residual_block(input_data,
512,
256,
512,
conv_trainable=False)
route_2 = input_data
input_data = common.convolutional(input_data, (3, 3, 512, 1024),
downsample=True,
conv_trainable=False)
for i in range(4):
input_data = common.residual_block(input_data,
1024,
512,
1024,
conv_trainable=False)
return route_1, route_2, input_data # [52,52,256], [26,26,512], [13,13,1024]
def cspdarknet53(input_data):
input_data = common.convolutional(input_data, (3, 3, 3, 32),
activate_type="mish")
input_data = common.convolutional(input_data, (3, 3, 32, 64),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 64, 64), activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 64, 64),
activate_type="mish")
for i in range(1):
input_data = common.residual_block(input_data, 64, 32, 64)
input_data = common.convolutional(input_data, (1, 1, 64, 64),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 128, 64),
activate_type="mish")
input_data = common.convolutional(input_data, (3, 3, 64, 128),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 128, 64), activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 128, 64),
activate_type="mish")
for i in range(2):
input_data = common.residual_block(input_data, 64, 64, 64)
input_data = common.convolutional(input_data, (1, 1, 64, 64),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 128, 128),
activate_type="mish")
input_data = common.convolutional(input_data, (3, 3, 128, 256),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 256, 128), activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 256, 128),
activate_type="mish")
for i in range(8):
input_data = common.residual_block(input_data, 128, 128, 128)
input_data = common.convolutional(input_data, (1, 1, 128, 128),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 256, 256),
activate_type="mish")
route_1 = input_data
input_data = common.convolutional(input_data, (3, 3, 256, 512),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 512, 256), activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 512, 256),
activate_type="mish")
for i in range(8):
input_data = common.residual_block(input_data, 256, 256, 256)
input_data = common.convolutional(input_data, (1, 1, 256, 256),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 512, 512),
activate_type="mish")
route_2 = input_data
input_data = common.convolutional(input_data, (3, 3, 512, 1024),
downsample=True,
activate_type="mish")
route = input_data
route = common.convolutional(route, (1, 1, 1024, 512),
activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 1024, 512),
activate_type="mish")
for i in range(4):
input_data = common.residual_block(input_data, 512, 512, 512)
input_data = common.convolutional(input_data, (1, 1, 512, 512),
activate_type="mish")
input_data = tf.concat([input_data, route], axis=-1)
input_data = common.convolutional(input_data, (1, 1, 1024, 1024),
activate_type="mish")
input_data = common.convolutional(input_data, (1, 1, 1024, 512))
input_data = common.convolutional(input_data, (3, 3, 512, 1024))
input_data = common.convolutional(input_data, (1, 1, 1024, 512))
input_data = tf.concat([
tf.nn.max_pool(input_data, ksize=13, padding='SAME', strides=1),
tf.nn.max_pool(input_data, ksize=9, padding='SAME', strides=1),
tf.nn.max_pool(input_data, ksize=5, padding='SAME', strides=1),
input_data
],
axis=-1)
input_data = common.convolutional(input_data, (1, 1, 2048, 512))
input_data = common.convolutional(input_data, (3, 3, 512, 1024))
input_data = common.convolutional(input_data, (1, 1, 1024, 512))
return route_1, route_2, input_data
