def load_network(args):
# load network
if args == 'vgg16':
net = VGG16()
#elif args == 'res50':
#elif args == 'res101':
#elif args == 'res152':
else:
raise NotImplementedError
return net
def SSD300(input_shape, num_classes=21):
# 300,300,3
input_tensor = Input(shape=input_shape)
img_size = (input_shape[1], input_shape[0])
# SSD结构,net字典
net = VGG16(input_tensor)
# -----------------------将提取到的主干特征进行处理---------------------------#
# 对conv4_3进行处理 38,38,512
net['conv4_3_norm'] = Normalize(20, name='conv4_3_norm')(net['conv4_3'])
num_priors = 4
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
net['conv4_3_norm_mbox_loc'] = Conv2D(num_priors * 4,
kernel_size=(3, 3),
padding='same',
name='conv4_3_norm_mbox_loc')(
net['conv4_3_norm'])
net['conv4_3_norm_mbox_loc_flat'] = Flatten(
name='conv4_3_norm_mbox_loc_flat')(net['conv4_3_norm_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
net['conv4_3_norm_mbox_conf'] = Conv2D(num_priors * num_classes,
kernel_size=(3, 3),
padding='same',
name='conv4_3_norm_mbox_conf')(
net['conv4_3_norm'])
net['conv4_3_norm_mbox_conf_flat'] = Flatten(
name='conv4_3_norm_mbox_conf_flat')(net['conv4_3_norm_mbox_conf'])
priorbox = PriorBox(img_size,
30.0,
max_size=60.0,
aspect_ratios=[2],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv4_3_norm_mbox_priorbox')
net['conv4_3_norm_mbox_priorbox'] = priorbox(net['conv4_3_norm'])
# 对fc7层进行处理
num_priors = 6
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
net['fc7_mbox_loc'] = Conv2D(num_priors * 4,
kernel_size=(3, 3),
padding='same',
name='fc7_mbox_loc')(net['fc7'])
net['fc7_mbox_loc_flat'] = Flatten(name='fc7_mbox_loc_flat')(
net['fc7_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
net['fc7_mbox_conf'] = Conv2D(num_priors * num_classes,
kernel_size=(3, 3),
padding='same',
name='fc7_mbox_conf')(net['fc7'])
net['fc7_mbox_conf_flat'] = Flatten(name='fc7_mbox_conf_flat')(
net['fc7_mbox_conf'])
priorbox = PriorBox(img_size,
60.0,
max_size=111.0,
aspect_ratios=[2, 3],
variances=[0.1, 0.1, 0.2, 0.2],
name='fc7_mbox_priorbox')
net['fc7_mbox_priorbox'] = priorbox(net['fc7'])
# 对conv6_2进行处理
num_priors = 6
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
x = Conv2D(num_priors * 4,
kernel_size=(3, 3),
padding='same',
name='conv6_2_mbox_loc')(net['conv6_2'])
net['conv6_2_mbox_loc'] = x
net['conv6_2_mbox_loc_flat'] = Flatten(name='conv6_2_mbox_loc_flat')(
net['conv6_2_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
x = Conv2D(num_priors * num_classes,
kernel_size=(3, 3),
padding='same',
name='conv6_2_mbox_conf')(net['conv6_2'])
net['conv6_2_mbox_conf'] = x
net['conv6_2_mbox_conf_flat'] = Flatten(name='conv6_2_mbox_conf_flat')(
net['conv6_2_mbox_conf'])
priorbox = PriorBox(img_size,
111.0,
max_size=162.0,
aspect_ratios=[2, 3],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv6_2_mbox_priorbox')
net['conv6_2_mbox_priorbox'] = priorbox(net['conv6_2'])
# 对conv7_2进行处理
num_priors = 6
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
x = Conv2D(num_priors * 4,
kernel_size=(3, 3),
padding='same',
name='conv7_2_mbox_loc')(net['conv7_2'])
net['conv7_2_mbox_loc'] = x
net['conv7_2_mbox_loc_flat'] = Flatten(name='conv7_2_mbox_loc_flat')(
net['conv7_2_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
x = Conv2D(num_priors * num_classes,
kernel_size=(3, 3),
padding='same',
name='conv7_2_mbox_conf')(net['conv7_2'])
net['conv7_2_mbox_conf'] = x
net['conv7_2_mbox_conf_flat'] = Flatten(name='conv7_2_mbox_conf_flat')(
net['conv7_2_mbox_conf'])
priorbox = PriorBox(img_size,
162.0,
max_size=213.0,
aspect_ratios=[2, 3],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv7_2_mbox_priorbox')
net['conv7_2_mbox_priorbox'] = priorbox(net['conv7_2'])
# 对conv8_2进行处理
num_priors = 4
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
x = Conv2D(num_priors * 4,
kernel_size=(3, 3),
padding='same',
name='conv8_2_mbox_loc')(net['conv8_2'])
net['conv8_2_mbox_loc'] = x
net['conv8_2_mbox_loc_flat'] = Flatten(name='conv8_2_mbox_loc_flat')(
net['conv8_2_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
x = Conv2D(num_priors * num_classes,
kernel_size=(3, 3),
padding='same',
name='conv8_2_mbox_conf')(net['conv8_2'])
net['conv8_2_mbox_conf'] = x
net['conv8_2_mbox_conf_flat'] = Flatten(name='conv8_2_mbox_conf_flat')(
net['conv8_2_mbox_conf'])
priorbox = PriorBox(img_size,
213.0,
max_size=264.0,
aspect_ratios=[2],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv8_2_mbox_priorbox')
net['conv8_2_mbox_priorbox'] = priorbox(net['conv8_2'])
# 对conv9_2进行处理
num_priors = 4
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
x = Conv2D(num_priors * 4,
kernel_size=(3, 3),
padding='same',
name='conv9_2_mbox_loc')(net['conv9_2'])
net['conv9_2_mbox_loc'] = x
net['conv9_2_mbox_loc_flat'] = Flatten(name='conv9_2_mbox_loc_flat')(
net['conv9_2_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
x = Conv2D(num_priors * num_classes,
kernel_size=(3, 3),
padding='same',
name='conv9_2_mbox_conf')(net['conv9_2'])
net['conv9_2_mbox_conf'] = x
net['conv9_2_mbox_conf_flat'] = Flatten(name='conv9_2_mbox_conf_flat')(
net['conv9_2_mbox_conf'])
priorbox = PriorBox(img_size,
264.0,
max_size=315.0,
aspect_ratios=[2],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv9_2_mbox_priorbox')
net['conv9_2_mbox_priorbox'] = priorbox(net['conv9_2'])
# 将所有结果进行堆叠
net['mbox_loc'] = concatenate([
net['conv4_3_norm_mbox_loc_flat'], net['fc7_mbox_loc_flat'],
net['conv6_2_mbox_loc_flat'], net['conv7_2_mbox_loc_flat'],
net['conv8_2_mbox_loc_flat'], net['conv9_2_mbox_loc_flat']
],
axis=1,
name='mbox_loc')
net['mbox_conf'] = concatenate([
net['conv4_3_norm_mbox_conf_flat'], net['fc7_mbox_conf_flat'],
net['conv6_2_mbox_conf_flat'], net['conv7_2_mbox_conf_flat'],
net['conv8_2_mbox_conf_flat'], net['conv9_2_mbox_conf_flat']
],
axis=1,
name='mbox_conf')
net['mbox_priorbox'] = concatenate([
net['conv4_3_norm_mbox_priorbox'], net['fc7_mbox_priorbox'],
net['conv6_2_mbox_priorbox'], net['conv7_2_mbox_priorbox'],
net['conv8_2_mbox_priorbox'], net['conv9_2_mbox_priorbox']
],
axis=1,
name='mbox_priorbox')
if hasattr(net['mbox_loc'], '_keras_shape'):
num_boxes = net['mbox_loc']._keras_shape[-1] // 4
elif hasattr(net['mbox_loc'], 'int_shape'):
num_boxes = K.int_shape(net['mbox_loc'])[-1] // 4
# 8732,4
net['mbox_loc'] = Reshape((num_boxes, 4),
name='mbox_loc_final')(net['mbox_loc'])
# 8732,21
net['mbox_conf'] = Reshape((num_boxes, num_classes),
name='mbox_conf_logits')(net['mbox_conf'])
net['mbox_conf'] = Activation('softmax',
name='mbox_conf_final')(net['mbox_conf'])
net['predictions'] = concatenate(
[net['mbox_loc'], net['mbox_conf'], net['mbox_priorbox']],
axis=2,
name='predictions')
print(net['predictions'])
model = Model(net['input'], net['predictions'])
return model
def SSD300(input_shape, num_classes=21, anchors_size=[30,60,111,162,213,264,315]):
#---------------------------------#
# 典型的输入大小为[300,300,3]
#---------------------------------#
input_tensor = Input(shape=input_shape)
# net变量里面包含了整个SSD的结构，通过层名可以找到对应的特征层
net = VGG16(input_tensor)
#-----------------------将提取到的主干特征进行处理---------------------------#
# 对conv4_3的通道进行l2标准化处理
# 38,38,512
net['conv4_3_norm'] = Normalize(20, name='conv4_3_norm')(net['conv4_3'])
num_priors = 4
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
net['conv4_3_norm_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same', name='conv4_3_norm_mbox_loc')(net['conv4_3_norm'])
net['conv4_3_norm_mbox_loc_flat'] = Flatten(name='conv4_3_norm_mbox_loc_flat')(net['conv4_3_norm_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
net['conv4_3_norm_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv4_3_norm_mbox_conf')(net['conv4_3_norm'])
net['conv4_3_norm_mbox_conf_flat'] = Flatten(name='conv4_3_norm_mbox_conf_flat')(net['conv4_3_norm_mbox_conf'])
priorbox = PriorBox(input_shape, anchors_size[0], max_size=anchors_size[1], aspect_ratios=[2],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv4_3_norm_mbox_priorbox')
net['conv4_3_norm_mbox_priorbox'] = priorbox(net['conv4_3_norm'])
# 对fc7层进行处理
# 19,19,1024
num_priors = 6
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
net['fc7_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3,3),padding='same',name='fc7_mbox_loc')(net['fc7'])
net['fc7_mbox_loc_flat'] = Flatten(name='fc7_mbox_loc_flat')(net['fc7_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
net['fc7_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3,3),padding='same',name='fc7_mbox_conf')(net['fc7'])
net['fc7_mbox_conf_flat'] = Flatten(name='fc7_mbox_conf_flat')(net['fc7_mbox_conf'])
priorbox = PriorBox(input_shape, anchors_size[1], max_size=anchors_size[2], aspect_ratios=[2, 3],
variances=[0.1, 0.1, 0.2, 0.2],
name='fc7_mbox_priorbox')
net['fc7_mbox_priorbox'] = priorbox(net['fc7'])
# 对conv6_2进行处理
# 10,10,512
num_priors = 6
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
x = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same',name='conv6_2_mbox_loc')(net['conv6_2'])
net['conv6_2_mbox_loc'] = x
net['conv6_2_mbox_loc_flat'] = Flatten(name='conv6_2_mbox_loc_flat')(net['conv6_2_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
x = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv6_2_mbox_conf')(net['conv6_2'])
net['conv6_2_mbox_conf'] = x
net['conv6_2_mbox_conf_flat'] = Flatten(name='conv6_2_mbox_conf_flat')(net['conv6_2_mbox_conf'])
priorbox = PriorBox(input_shape, anchors_size[2], max_size=anchors_size[3], aspect_ratios=[2, 3],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv6_2_mbox_priorbox')
net['conv6_2_mbox_priorbox'] = priorbox(net['conv6_2'])
# 对conv7_2进行处理
# 5,5,256
num_priors = 6
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
x = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same',name='conv7_2_mbox_loc')(net['conv7_2'])
net['conv7_2_mbox_loc'] = x
net['conv7_2_mbox_loc_flat'] = Flatten(name='conv7_2_mbox_loc_flat')(net['conv7_2_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
x = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv7_2_mbox_conf')(net['conv7_2'])
net['conv7_2_mbox_conf'] = x
net['conv7_2_mbox_conf_flat'] = Flatten(name='conv7_2_mbox_conf_flat')(net['conv7_2_mbox_conf'])
priorbox = PriorBox(input_shape, anchors_size[3], max_size=anchors_size[4], aspect_ratios=[2, 3],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv7_2_mbox_priorbox')
net['conv7_2_mbox_priorbox'] = priorbox(net['conv7_2'])
# 对conv8_2进行处理
# 3,3,256
num_priors = 4
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
x = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same',name='conv8_2_mbox_loc')(net['conv8_2'])
net['conv8_2_mbox_loc'] = x
net['conv8_2_mbox_loc_flat'] = Flatten(name='conv8_2_mbox_loc_flat')(net['conv8_2_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
x = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv8_2_mbox_conf')(net['conv8_2'])
net['conv8_2_mbox_conf'] = x
net['conv8_2_mbox_conf_flat'] = Flatten(name='conv8_2_mbox_conf_flat')(net['conv8_2_mbox_conf'])
priorbox = PriorBox(input_shape, anchors_size[4], max_size=anchors_size[5], aspect_ratios=[2],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv8_2_mbox_priorbox')
net['conv8_2_mbox_priorbox'] = priorbox(net['conv8_2'])
# 对conv9_2进行处理
# 1,1,256
num_priors = 4
# 预测框的处理
# num_priors表示每个网格点先验框的数量，4是x,y,h,w的调整
x = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same',name='conv9_2_mbox_loc')(net['conv9_2'])
net['conv9_2_mbox_loc'] = x
net['conv9_2_mbox_loc_flat'] = Flatten(name='conv9_2_mbox_loc_flat')(net['conv9_2_mbox_loc'])
# num_priors表示每个网格点先验框的数量，num_classes是所分的类
x = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv9_2_mbox_conf')(net['conv9_2'])
net['conv9_2_mbox_conf'] = x
net['conv9_2_mbox_conf_flat'] = Flatten(name='conv9_2_mbox_conf_flat')(net['conv9_2_mbox_conf'])
priorbox = PriorBox(input_shape, anchors_size[5], max_size=anchors_size[6], aspect_ratios=[2],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv9_2_mbox_priorbox')
net['conv9_2_mbox_priorbox'] = priorbox(net['conv9_2'])
# 将所有结果进行堆叠
net['mbox_loc'] = concatenate([net['conv4_3_norm_mbox_loc_flat'],
net['fc7_mbox_loc_flat'],
net['conv6_2_mbox_loc_flat'],
net['conv7_2_mbox_loc_flat'],
net['conv8_2_mbox_loc_flat'],
net['conv9_2_mbox_loc_flat']],
axis=1, name='mbox_loc')
net['mbox_conf'] = concatenate([net['conv4_3_norm_mbox_conf_flat'],
net['fc7_mbox_conf_flat'],
net['conv6_2_mbox_conf_flat'],
net['conv7_2_mbox_conf_flat'],
net['conv8_2_mbox_conf_flat'],
net['conv9_2_mbox_conf_flat']],
axis=1, name='mbox_conf')
net['mbox_priorbox'] = concatenate([net['conv4_3_norm_mbox_priorbox'],
net['fc7_mbox_priorbox'],
net['conv6_2_mbox_priorbox'],
net['conv7_2_mbox_priorbox'],
net['conv8_2_mbox_priorbox'],
net['conv9_2_mbox_priorbox']],
axis=1, name='mbox_priorbox')
# 8732,4
net['mbox_loc'] = Reshape((-1, 4),name='mbox_loc_final')(net['mbox_loc'])
# 8732,21
net['mbox_conf'] = Reshape((-1, num_classes),name='mbox_conf_logits')(net['mbox_conf'])
# 8732,8
net['mbox_conf'] = Activation('softmax',name='mbox_conf_final')(net['mbox_conf'])
# 8732,33
net['predictions'] = concatenate([net['mbox_loc'],
net['mbox_conf'],
net['mbox_priorbox']],
axis=2, name='predictions')
model = Model(net['input'], net['predictions'])
return model
def SSD300(input_shape, num_classes=21):
# 300,300,3
input_tensor = Input(shape=input_shape)
img_size = (input_shape[1], input_shape[0])
# SSD architecture,net
net = VGG16(input_tensor)
#-----------------------Six feature layers Processing---------------------------#
# Process conv4_3 (38,38,512)
net['conv4_3_norm'] = Normalize(20, name='conv4_3_norm')(net['conv4_3'])
num_priors = 4
# Process the Prediction of bounding box
# num_priors means how many priors in cell of this feature layer ，4 is x,y,h,w, positio of bounding box.
net['conv4_3_norm_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same', name='conv4_3_norm_mbox_loc')(net['conv4_3_norm'])
net['conv4_3_norm_mbox_loc_flat'] = Flatten(name='conv4_3_norm_mbox_loc_flat')(net['conv4_3_norm_mbox_loc'])
# num_priors means how many priors in cell of this feature layer，num_classes means object classs
net['conv4_3_norm_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv4_3_norm_mbox_conf')(net['conv4_3_norm'])
net['conv4_3_norm_mbox_conf_flat'] = Flatten(name='conv4_3_norm_mbox_conf_flat')(net['conv4_3_norm_mbox_conf'])
priorbox = PriorBox(img_size, 30.0,max_size = 60.0, aspect_ratios=[2],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv4_3_norm_mbox_priorbox')
net['conv4_3_norm_mbox_priorbox'] = priorbox(net['conv4_3_norm'])
# Process fc7
num_priors = 6
# Process the Prediction of bounding box
# num_priors means how many priors in cell of this feature layer ，4 is x,y,h,w, positio of bounding box.
net['fc7_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3,3),padding='same',name='fc7_mbox_loc')(net['fc7'])
net['fc7_mbox_loc_flat'] = Flatten(name='fc7_mbox_loc_flat')(net['fc7_mbox_loc'])
# num_priors means how many priors in cell of this feature layer，num_classes means object classs
net['fc7_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3,3),padding='same',name='fc7_mbox_conf')(net['fc7'])
net['fc7_mbox_conf_flat'] = Flatten(name='fc7_mbox_conf_flat')(net['fc7_mbox_conf'])
priorbox = PriorBox(img_size, 60.0, max_size=111.0, aspect_ratios=[2, 3],
variances=[0.1, 0.1, 0.2, 0.2],
name='fc7_mbox_priorbox')
net['fc7_mbox_priorbox'] = priorbox(net['fc7'])
# Process conv6_2
num_priors = 6
# Process the Prediction of bounding box
# num_priors means how many priors in cell of this feature layer ，4 is x,y,h,w, positio of bounding box.
x = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same',name='conv6_2_mbox_loc')(net['conv6_2'])
net['conv6_2_mbox_loc'] = x
net['conv6_2_mbox_loc_flat'] = Flatten(name='conv6_2_mbox_loc_flat')(net['conv6_2_mbox_loc'])
# num_priors means how many priors in cell of this feature layer，num_classes means object classs
x = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv6_2_mbox_conf')(net['conv6_2'])
net['conv6_2_mbox_conf'] = x
net['conv6_2_mbox_conf_flat'] = Flatten(name='conv6_2_mbox_conf_flat')(net['conv6_2_mbox_conf'])
priorbox = PriorBox(img_size, 111.0, max_size=162.0, aspect_ratios=[2, 3],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv6_2_mbox_priorbox')
net['conv6_2_mbox_priorbox'] = priorbox(net['conv6_2'])
# Process conv7_2
num_priors = 6
# Process the Prediction of bounding box
# num_priors means how many priors in cell of this feature layer ，4 is x,y,h,w, positio of bounding box.
x = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same',name='conv7_2_mbox_loc')(net['conv7_2'])
net['conv7_2_mbox_loc'] = x
net['conv7_2_mbox_loc_flat'] = Flatten(name='conv7_2_mbox_loc_flat')(net['conv7_2_mbox_loc'])
# num_priors means how many priors in cell of this feature layer，num_classes means object classs
x = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv7_2_mbox_conf')(net['conv7_2'])
net['conv7_2_mbox_conf'] = x
net['conv7_2_mbox_conf_flat'] = Flatten(name='conv7_2_mbox_conf_flat')(net['conv7_2_mbox_conf'])
priorbox = PriorBox(img_size, 162.0, max_size=213.0, aspect_ratios=[2, 3],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv7_2_mbox_priorbox')
net['conv7_2_mbox_priorbox'] = priorbox(net['conv7_2'])
# Process conv8_2
num_priors = 4
# Process the Prediction of bounding box
# num_priors means how many priors in cell of this feature layer ，4 is x,y,h,w, positio of bounding box.
x = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same',name='conv8_2_mbox_loc')(net['conv8_2'])
net['conv8_2_mbox_loc'] = x
net['conv8_2_mbox_loc_flat'] = Flatten(name='conv8_2_mbox_loc_flat')(net['conv8_2_mbox_loc'])
# num_priors means how many priors in cell of this feature layer，num_classes means object classs
x = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv8_2_mbox_conf')(net['conv8_2'])
net['conv8_2_mbox_conf'] = x
net['conv8_2_mbox_conf_flat'] = Flatten(name='conv8_2_mbox_conf_flat')(net['conv8_2_mbox_conf'])
priorbox = PriorBox(img_size, 213.0, max_size=264.0, aspect_ratios=[2],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv8_2_mbox_priorbox')
net['conv8_2_mbox_priorbox'] = priorbox(net['conv8_2'])
# Process conv9_2
num_priors = 4
# Process the Prediction of bounding box
# num_priors means how many priors in cell of this feature layer ，4 is x,y,h,w, positio of bounding box.
x = Conv2D(num_priors * 4, kernel_size=(3,3), padding='same',name='conv9_2_mbox_loc')(net['conv9_2'])
net['conv9_2_mbox_loc'] = x
net['conv9_2_mbox_loc_flat'] = Flatten(name='conv9_2_mbox_loc_flat')(net['conv9_2_mbox_loc'])
# num_priors means how many priors in cell of this feature layer，num_classes means object classs
x = Conv2D(num_priors * num_classes, kernel_size=(3,3), padding='same',name='conv9_2_mbox_conf')(net['conv9_2'])
net['conv9_2_mbox_conf'] = x
net['conv9_2_mbox_conf_flat'] = Flatten(name='conv9_2_mbox_conf_flat')(net['conv9_2_mbox_conf'])
priorbox = PriorBox(img_size, 264.0, max_size=315.0, aspect_ratios=[2],
variances=[0.1, 0.1, 0.2, 0.2],
name='conv9_2_mbox_priorbox')
net['conv9_2_mbox_priorbox'] = priorbox(net['conv9_2'])
# concatenate location and conffident in all feature layers
net['mbox_loc'] = concatenate([net['conv4_3_norm_mbox_loc_flat'],
net['fc7_mbox_loc_flat'],
net['conv6_2_mbox_loc_flat'],
net['conv7_2_mbox_loc_flat'],
net['conv8_2_mbox_loc_flat'],
net['conv9_2_mbox_loc_flat']],
axis=1, name='mbox_loc')
net['mbox_conf'] = concatenate([net['conv4_3_norm_mbox_conf_flat'],
net['fc7_mbox_conf_flat'],
net['conv6_2_mbox_conf_flat'],
net['conv7_2_mbox_conf_flat'],
net['conv8_2_mbox_conf_flat'],
net['conv9_2_mbox_conf_flat']],
axis=1, name='mbox_conf')
net['mbox_priorbox'] = concatenate([net['conv4_3_norm_mbox_priorbox'],
net['fc7_mbox_priorbox'],
net['conv6_2_mbox_priorbox'],
net['conv7_2_mbox_priorbox'],
net['conv8_2_mbox_priorbox'],
net['conv9_2_mbox_priorbox']],
axis=1, name='mbox_priorbox')
if hasattr(net['mbox_loc'], '_keras_shape'):
num_boxes = net['mbox_loc']._keras_shape[-1] // 4
elif hasattr(net['mbox_loc'], 'int_shape'):
num_boxes = K.int_shape(net['mbox_loc'])[-1] // 4
# 8732,4
net['mbox_loc'] = Reshape((num_boxes, 4),name='mbox_loc_final')(net['mbox_loc'])
# 8732,21
net['mbox_conf'] = Reshape((num_boxes, num_classes),name='mbox_conf_logits')(net['mbox_conf'])
net['mbox_conf'] = Activation('softmax',name='mbox_conf_final')(net['mbox_conf'])
net['predictions'] = concatenate([net['mbox_loc'],
net['mbox_conf'],
net['mbox_priorbox']],
axis=2, name='predictions')
print(net['predictions'])
model = Model(net['input'], net['predictions'])
return model
def SSD300(input_shape, num_classes):
input_tensor = Input(shape=input_shape)
img_size = (input_shape[1], input_shape[0]) # last channel
net = VGG16(input_tensor)
# 38x38x512
net['conv4_3_norm'] = Normalize(20, name='conv4_3_norm') (net['conv4_3'])
num_priors = 4
# 38x38x16 first featureMap
net['conv4_3_norm_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3, 3), padding='same', name='conv4_3_norm_mbox_loc') (net['conv4_3_norm'])
net['conv4_3_norm_mbox_loc_flat'] = Flatten(name='conv4_3_norm_mbox_loc_flat') (net['conv4_3_norm_mbox_loc'])
# [38, 38, num_classes * num_priors] first featureMap
net['conv4_3_norm_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3, 3), padding='same', name='conv4_3_norm_mbox_conf') (net['conv4_3_norm'])
net['conv4_3_norm_mbox_conf_flat'] = Flatten(name='onv4_3_norm_mbox_conf_flat') (net['conv4_3_norm_mbox_conf'])
priorbox = PriorBox(img_size, 30.0, max_size=60.0, aspect_ratios=[2], variances=[0.1, 0.1, 0.2, 0.2], name='conv4_3_norm_mbox_priorbox')
# 38 * 38 * 4 = 5576
net['conv4_3_norm_mbox_priorbox'] = priorbox.call(net['conv4_3_norm'])
num_priors = 6
# [19, 19, 24]
net['fc7_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3, 3), padding='same', name='fc7_mbox_loc') (net['fc7'])
net['fc7_mbox_loc_flat'] = Flatten(name='fc7_mbox_loc_flat') (net['fc7_mbox_loc'])
# [19, 19, num_priors * num_classes] second featureMap
net['fc7_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3, 3), padding='same', name='fc7_mbox_conf') (net['fc7'])
net['fc7_mbox_conf_flat'] = Flatten(name='fc7_mbox_conf_flat') (net['fc7_mbox_conf'])
priorbox = PriorBox(img_size, 60.0, max_size=111.0, aspect_ratios=[2, 3], variances=[0.1, 0.1, 0.2, 0.2], name='fc7_mbox_priorbox')
# 19 * 19 * 6 = 2166
net['fc7_mbox_priorbox'] = priorbox(net['fc7'])
num_priors = 6
# [10, 10, 24]
net['conv6_2_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3, 3), padding='same', name='conv6_2_mbox_loc') (net['conv6_2'])
net['conv6_2_mbox_loc_flat'] = Flatten(name='conv6_2_mbox_loc_flat') (net['conv6_2_mbox_loc'])
net['conv6_2_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3, 3), padding='same', name='conv6_2_mbox_conf') (net['conv6_2'])
net['conv6_2_mbox_conf_flat'] = Flatten(name='conv6_2_mbox_conf_flat') (net['conv6_2_mbox_conf'])
priorbox = PriorBox(img_size, 111.0, max_size=162.0, aspect_ratios=[2, 3], variances=[0.1, 0.1, 0.2, 0.2], name='conv6_2_mbox_priorbox')
# 10x10x6 = 600
net['conv6_2_mbox_priorbox'] = priorbox(net['conv6_2'])
num_priors = 6
# [5, 5, 24]
net['conv7_2_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3, 3), padding='same', name='conv7_2_mbox_loc') (net['conv7_2'])
net['conv7_2_mbox_loc_flat'] = Flatten(name='conv7_2_mbox_loc_flat') (net['conv7_2_mbox_loc'])
net['conv7_2_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3, 3), padding='same', name='conv7_2_mbox_conf') (net['conv7_2'])
net['conv7_2_mbox_conf_flat'] = Flatten(name='conv7_2_mbox_conf_flat') (net['conv7_2_mbox_conf'])
priorbox = PriorBox(img_size, 162.0, max_size=213.0, aspect_ratios=[2, 3], variances=[0.1, 0.1, 0.2, 0.2], name='conv7_2_mbox_priorbox')
# 5x5x6 = 150
net['conv7_2_mbox_priorbox'] = priorbox(net['conv7_2'])
num_priors = 4
net['conv8_2_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3, 3), padding='same', name='conv8_2_mbox_loc') (net['conv8_2'])
net['conv8_2_mbox_loc_flat'] = Flatten(name='conv8_2_mbox_loc_flat') (net['conv8_2_mbox_loc'])
net['conv8_2_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3, 3), padding='same', name='conv8_2_mbox_conf') (net['conv8_2'])
net['conv8_2_mbox_conf_flat'] = Flatten(name='conv8_2_mbox_conf_flat') (net['conv8_2_mbox_conf'])
# 3x3x4 = 36
priorbox = PriorBox(img_size, 213.0, max_size=264.0, aspect_ratios=[2], variances=[0.1, 0.1, 0.2, 0.2], name='conv8_2_mbox_priorbox')
net['conv8_2_mbox_priorbox'] = priorbox(net['conv8_2'])
num_priors = 4
net['conv9_2_mbox_loc'] = Conv2D(num_priors * 4, kernel_size=(3, 3), padding='same', name='conv9_2_mbox_loc') (net['conv9_2'])
net['conv9_2_mbox_loc_flat'] = Flatten(name='conv9_2_mbox_loc_flat') (net['conv9_2_mbox_loc'])
net['conv9_2_mbox_conf'] = Conv2D(num_priors * num_classes, kernel_size=(3, 3), padding='same', name='conv9_2_mbox_conf') (net['conv9_2'])
net['conv9_2_mbox_conf_flat'] = Flatten(name='conv9_2_mbox_conf_flat') (net['conv9_2_mbox_conf'])
# 1x1x4
priorbox = PriorBox(img_size, 264.0, max_size=315.0, aspect_ratios=[2], variances=[0.1, 0.1, 0.2, 0.2], name='conv9_2_mbox_priorbox')
net['conv9_2_mbox_priorbox'] = priorbox(net['conv9_2'])
# a = net['conv4_3_norm_mbox_loc_flat']
# b = net['fc7_mbox_loc_flat']
# d = net['conv6_2_mbox_loc_flat']
# e = net['conv7_2_mbox_loc_flat']
# f = net['conv8_2_mbox_loc_flat']
# g = net['conv9_2_mbox_loc_flat']
# v = net['mbox_loc']
net['mbox_loc'] = Concatenate(axis=1, name='mbox_loc') ([
net['conv4_3_norm_mbox_loc_flat'],
net['fc7_mbox_loc_flat'],
net['conv6_2_mbox_loc_flat'],
net['conv7_2_mbox_loc_flat'],
net['conv8_2_mbox_loc_flat'],
net['conv9_2_mbox_loc_flat']
])
net['mbox_conf'] = Concatenate(axis=1, name='mbox_conf') ([
net['conv4_3_norm_mbox_conf_flat'],
net['fc7_mbox_conf_flat'],
net['conv6_2_mbox_conf_flat'],
net['conv7_2_mbox_conf_flat'],
net['conv8_2_mbox_conf_flat'],
net['conv9_2_mbox_conf_flat']
])
net['mbox_priorbox'] = Concatenate(axis=1, name='mbox_priorbox') ([
net['conv4_3_norm_mbox_priorbox'],
net['fc7_mbox_priorbox'],
net['conv6_2_mbox_priorbox'],
net['conv7_2_mbox_priorbox'],
net['conv8_2_mbox_priorbox'],
net['conv9_2_mbox_priorbox']
])
# net['mbox_loc'] = tf.reshape((-1, 4), name='mbox_loc_final') (net['mbox_loc'])
# 最终8732个预测框, 回归问题
net['mbox_loc'] = tf.keras.layers.Reshape((-1, 4), name='mbox_loc_final') (net['mbox_loc'])
# finally [8732, num_classes]
net['mbox_conf'] = tf.keras.layers.Reshape((-1, num_classes), name='mbox_conf_logits') (net['mbox_conf'])
net['mbox_conf'] = tf.keras.layers.Activation('softmax', name='mbox_conf_final') (net['mbox_conf'])
net['predictions'] = Concatenate(axis=2, name='predictions') ([
net['mbox_loc'],
net['mbox_conf'],
net['mbox_priorbox']
])
model = tf.keras.Model(net['input'], net['predictions']) # 闭包
return model