代码来自b站/csdn 大佬：Bubbliiiing
一.网络结构
1.darknet.py
from functools import wraps
from tensorflow.keras.initializers import RandomNormal
from tensorflow.keras.layers import (Add, BatchNormalization, Conv2D, LeakyReLU,
ZeroPadding2D)
from tensorflow.keras.regularizers import l2
from utils.utils import compose
#------------------------------------------------------#
# 单次卷积
# DarknetConv2D
#------------------------------------------------------#
@wraps(Conv2D)
def DarknetConv2D(*args, **kwargs):
darknet_conv_kwargs = {'kernel_initializer' : RandomNormal(stddev=0.02), 'kernel_regularizer': l2(5e-4)}
darknet_conv_kwargs['padding'] = 'valid' if kwargs.get('strides')==(2, 2) else 'same'
darknet_conv_kwargs.update(kwargs)
return Conv2D(*args, **darknet_conv_kwargs)
#---------------------------------------------------#
# 卷积块 -> 卷积 + 标准化 + 激活函数
# DarknetConv2D + BatchNormalization + LeakyReLU
#---------------------------------------------------#
def DarknetConv2D_BN_Leaky(*args, **kwargs):
no_bias_kwargs = {'use_bias': False}
no_bias_kwargs.update(kwargs)
return compose(
DarknetConv2D(*args, **no_bias_kwargs),
BatchNormalization(),
LeakyReLU(alpha=0.1))
#---------------------------------------------------------------------#
# 残差结构
# 首先利用ZeroPadding2D和一个步长为2x2的卷积块进行高和宽的压缩
# 然后对num_blocks进行循环，循环内部是残差结构。
#---------------------------------------------------------------------#
def resblock_body(x, num_filters, num_blocks):
x = ZeroPadding2D(((1,0),(1,0)))(x)
x = DarknetConv2D_BN_Leaky(num_filters, (3,3), strides=(2,2))(x)
for i in range(num_blocks):
y = DarknetConv2D_BN_Leaky(num_filters//2, (1,1))(x)
y = DarknetConv2D_BN_Leaky(num_filters, (3,3))(y)
x = Add()([x,y])
return x
#---------------------------------------------------#
# darknet53 的主体部分
# 输入为一张416x416x3的图片
# 输出为三个有效特征层
#---------------------------------------------------#
def darknet_body(x):
# 416,416,3 -> 416,416,32
x = DarknetConv2D_BN_Leaky(32, (3,3))(x)
# 416,416,32 -> 208,208,64
x = resblock_body(x, 64, 1)
# 208,208,64 -> 104,104,128
x = resblock_body(x, 128, 2)
# 104,104,128 -> 52,52,256
x = resblock_body(x, 256, 8)
feat1 = x
# 52,52,256 -> 26,26,512
x = resblock_body(x, 512, 8)
feat2 = x
# 26,26,512 -> 13,13,1024
x = resblock_body(x, 1024, 4)
feat3 = x
return feat1, feat2, feat3
@wraps(Conv2D)是装饰器，传入的是一个已经有的函数a，返回的是一个新的函数b。在函数Conv2D的基础上，我们可以修改一些Conv2D中已经存在的参数，然后返回一个新的Conv2D供我们使用。可以参考这篇博客理解https://blog.csdn.net/weixin_38495031/article/details/105357313
'kernel_initializer' : RandomNormal(stddev=0.02), 'kernel_regularizer': l2(5e-4)
这句是采用正态分布初始化卷积核，标准差为0.02.采用L2正则化，lambda为5e-4。
'padding'] = 'valid' if kwargs.get('strides')==(2, 2) else 'same'
这句是如果卷积的步长为2，2，则padding的填充方式为valid，否则为same。
darknet_conv_kwargs.update(kwargs)
这句是更新参数,通过装饰器我们修改了conv2d的参数。
def DarknetConv2D_BN_Leaky(*args, **kwargs):
no_bias_kwargs = {'use_bias': False}
no_bias_kwargs.update(kwargs)
return compose(
DarknetConv2D(*args, **no_bias_kwargs),
BatchNormalization(),
LeakyReLU(alpha=0.1))
DarknetConv2D_BN_Leaky是一个融合了卷积，bn和leakyRelu激活的三层结构。‘use_bias’: False表示取消卷积的bias。是因为我们后面的BatchNorm层会加上自己的bias，所以我们的卷积层的（如果有）bias就是多余的了。
x = ZeroPadding2D(((1,0),(1,0)))(x)
ZeroPadding2D，传入的参数如果是一个二维的tuple，((top_pad, bottom_pad), (left_pad, right_pad))，它表示在上下左右分别补多少层零。这句是对输入的图像左侧和上侧填充一行0，如下
[[0. 0. 0. 0.]
[0. 1. 2. 3.]
[0. 4. 5. 6.]
[0. 7. 8. 9.]]
2.yolo.py
from tensorflow.keras.layers import Concatenate, Input, Lambda, UpSampling2D
from tensorflow.keras.models import Model
from utils.utils import compose
from nets.darknet import DarknetConv2D, DarknetConv2D_BN_Leaky, darknet_body
from nets.yolo_training import yolo_loss
#---------------------------------------------------#
# 特征层->最后的输出
#---------------------------------------------------#
def make_five_conv(x, num_filters):
x = DarknetConv2D_BN_Leaky(num_filters, (1,1))(x)
x = DarknetConv2D_BN_Leaky(num_filters*2, (3,3))(x)
x = DarknetConv2D_BN_Leaky(num_filters, (1,1))(x)
x = DarknetConv2D_BN_Leaky(num_filters*2, (3,3))(x)
x = DarknetConv2D_BN_Leaky(num_filters, (1,1))(x)
return x
def make_yolo_head(x, num_filters, out_filters):
y = DarknetConv2D_BN_Leaky(num_filters*2, (3,3))(x)
# 255->3, 85->3, 4 + 1 + 80
y = DarknetConv2D(out_filters, (1,1))(y)
return y
#---------------------------------------------------#
# FPN网络的构建，并且获得预测结果
#---------------------------------------------------#
def yolo_body(input_shape, anchors_mask, num_classes):
inputs = Input(input_shape)
#---------------------------------------------------#
# 生成darknet53的主干模型
# 获得三个有效特征层，他们的shape分别是：
# C3 为 52,52,256
# C4 为 26,26,512
# C5 为 13,13,1024
#---------------------------------------------------#
C3, C4, C5 = darknet_body(inputs)
#---------------------------------------------------#
# 第一个特征层
# y1=(batch_size,13,13,3,85)
#---------------------------------------------------#
# 13,13,1024 -> 13,13,512 -> 13,13,1024 -> 13,13,512 -> 13,13,1024 -> 13,13,512
x = make_five_conv(C5, 512)
P5 = make_yolo_head(x, 512, len(anchors_mask[0]) * (num_classes+5))
# 13,13,512 -> 13,13,256 -> 26,26,256
x = compose(DarknetConv2D_BN_Leaky(256, (1,1)), UpSampling2D(2))(x)
# 26,26,256 + 26,26,512 -> 26,26,768
x = Concatenate()([x, C4])
#---------------------------------------------------#
# 第二个特征层
# y2=(batch_size,26,26,3,85)
#---------------------------------------------------#
# 26,26,768 -> 26,26,256 -> 26,26,512 -> 26,26,256 -> 26,26,512 -> 26,26,256
x = make_five_conv(x, 256)
P4 = make_yolo_head(x, 256, len(anchors_mask[1]) * (num_classes+5))
# 26,26,256 -> 26,26,128 -> 52,52,128
x = compose(DarknetConv2D_BN_Leaky(128, (1,1)), UpSampling2D(2))(x)
# 52,52,128 + 52,52,256 -> 52,52,384
x = Concatenate()([x, C3])
#---------------------------------------------------#
# 第三个特征层
# y3=(batch_size,52,52,3,85)
#---------------------------------------------------#
# 52,52,384 -> 52,52,128 -> 52,52,256 -> 52,52,128 -> 52,52,256 -> 52,52,128
x = make_five_conv(x, 128)
P3 = make_yolo_head(x, 128, len(anchors_mask[2]) * (num_classes+5))
return Model(inputs, [P5, P4, P3])
def get_train_model(model_body, input_shape, num_classes, anchors, anchors_mask):
y_true = [Input(shape = (input_shape[0] // {0:32, 1:16, 2:8}[l], input_shape[1] // {0:32, 1:16, 2:8}[l], \
len(anchors_mask[l]), num_classes + 5)) for l in range(len(anchors_mask))]
model_loss = Lambda(
yolo_loss,
output_shape = (1, ),
name = 'yolo_loss',
arguments = {'input_shape' : input_shape, 'anchors' : anchors, 'anchors_mask' : anchors_mask, 'num_classes' : num_classes}
)([*model_body.output, *y_true])
model = Model([model_body.input, *y_true], model_loss)
return model
inputs = Input(input_shape)
Input（）`用于实例化Keras张量。
make_five_conv是图中的
make_yolo_head是图中的
anchors_mask[0] anchors_mask[1] anchors_mask[2] 分别是[6, 7, 8]，[3, 4, 5]， [0, 1, 2]，代表每个三个特征层每层的每个grid cell都生成三个先验框。用这样一个9个数的数组来分别对应如下聚类得到的九个先验框大小。
其余代码都很容易理解。
二.先验框产生和调整代码
import tensorflow as tf
from tensorflow.keras import backend as K
#---------------------------------------------------#
# 对box进行调整，使其符合真实图片的样子
#---------------------------------------------------#
def yolo_correct_boxes(box_xy, box_wh, input_shape, image_shape, letterbox_image):
#-----------------------------------------------------------------#
# 把y轴放前面是因为方便预测框和图像的宽高进行相乘
#-----------------------------------------------------------------#
box_yx = box_xy[..., ::-1]
box_hw = box_wh[..., ::-1]
input_shape = K.cast(input_shape, K.dtype(box_yx))
image_shape = K.cast(image_shape, K.dtype(box_yx))
if letterbox_image:
#-----------------------------------------------------------------#
# 这里求出来的offset是图像有效区域相对于图像左上角的偏移情况
# new_shape指的是宽高缩放情况
#-----------------------------------------------------------------#
new_shape = K.round(image_shape * K.min(input_shape/image_shape))
offset = (input_shape - new_shape)/2./input_shape
scale = input_shape/new_shape
box_yx = (box_yx - offset) * scale
box_hw *= scale
box_mins = box_yx - (box_hw / 2.)
box_maxes = box_yx + (box_hw / 2.)
boxes = K.concatenate([box_mins[..., 0:1], box_mins[..., 1:2], box_maxes[..., 0:1], box_maxes[..., 1:2]])
boxes *= K.concatenate([image_shape, image_shape])
return boxes
#---------------------------------------------------#
# 将预测值的每个特征层调成真实值
#---------------------------------------------------#
def get_anchors_and_decode(feats, anchors, num_classes, input_shape, calc_loss=False):
num_anchors = len(anchors)
#------------------------------------------#
# grid_shape指的是特征层的高和宽
#------------------------------------------#
grid_shape = K.shape(feats)[1:3]
#--------------------------------------------------------------------#
# 获得各个特征点的坐标信息。生成的shape为(13, 13, num_anchors, 2)
#--------------------------------------------------------------------#
grid_x = K.tile(K.reshape(K.arange(0, stop=grid_shape[1]), [1, -1, 1, 1]), [grid_shape[0], 1, num_anchors, 1])
grid_y = K.tile(K.reshape(K.arange(0, stop=grid_shape[0]), [-1, 1, 1, 1]), [1, grid_shape[1], num_anchors, 1])
grid = K.cast(K.concatenate([grid_x, grid_y]), K.dtype(feats))
#---------------------------------------------------------------#
# 将先验框进行拓展，生成的shape为(13, 13, num_anchors, 2)
#---------------------------------------------------------------#
anchors_tensor = K.reshape(K.constant(anchors), [1, 1, num_anchors, 2])
anchors_tensor = K.tile(anchors_tensor, [grid_shape[0], grid_shape[1], 1, 1])
#---------------------------------------------------#
# 将预测结果调整成(batch_size,13,13,3,85)
# 85可拆分成4 + 1 + 80
# 4代表的是中心宽高的调整参数
# 1代表的是框的置信度
# 80代表的是种类的置信度
#---------------------------------------------------#
feats = K.reshape(feats, [-1, grid_shape[0], grid_shape[1], num_anchors, num_classes + 5])
#------------------------------------------#
# 对先验框进行解码，并进行归一化
#------------------------------------------#
box_xy = (K.sigmoid(feats[..., :2]) + grid) / K.cast(grid_shape[::-1], K.dtype(feats))
box_wh = K.exp(feats[..., 2:4]) * anchors_tensor / K.cast(input_shape[::-1], K.dtype(feats))
#------------------------------------------#
# 获得预测框的置信度
#------------------------------------------#
box_confidence = K.sigmoid(feats[..., 4:5])
box_class_probs = K.sigmoid(feats[..., 5:])
#---------------------------------------------------------------------#
# 在计算loss的时候返回grid, feats, box_xy, box_wh
# 在预测的时候返回box_xy, box_wh, box_confidence, box_class_probs
#---------------------------------------------------------------------#
if calc_loss == True:
return grid, feats, box_xy, box_wh
return box_xy, box_wh, box_confidence, box_class_probs
#---------------------------------------------------#
# 图片预测
#---------------------------------------------------#
def DecodeBox(outputs,
anchors,
num_classes,
input_shape,
#-----------------------------------------------------------#
# 13x13的特征层对应的anchor是[116,90],[156,198],[373,326]
# 26x26的特征层对应的anchor是[30,61],[62,45],[59,119]
# 52x52的特征层对应的anchor是[10,13],[16,30],[33,23]
#-----------------------------------------------------------#
anchor_mask = [[6, 7, 8], [3, 4, 5], [0, 1, 2]],
max_boxes = 100,
confidence = 0.5,
nms_iou = 0.3,
letterbox_image = True):
image_shape = K.reshape(outputs[-1],[-1])
box_xy = []
box_wh = []
box_confidence = []
box_class_probs = []
for i in range(len(anchor_mask)):
sub_box_xy, sub_box_wh, sub_box_confidence, sub_box_class_probs = \
get_anchors_and_decode(outputs[i], anchors[anchor_mask[i]], num_classes, input_shape)
box_xy.append(K.reshape(sub_box_xy, [-1, 2]))
box_wh.append(K.reshape(sub_box_wh, [-1, 2]))
box_confidence.append(K.reshape(sub_box_confidence, [-1, 1]))
box_class_probs.append(K.reshape(sub_box_class_probs, [-1, num_classes]))
box_xy = K.concatenate(box_xy, axis = 0)
box_wh = K.concatenate(box_wh, axis = 0)
box_confidence = K.concatenate(box_confidence, axis = 0)
box_class_probs = K.concatenate(box_class_probs, axis = 0)
#------------------------------------------------------------------------------------------------------------#
# 在图像传入网络预测前会进行letterbox_image给图像周围添加灰条，因此生成的box_xy, box_wh是相对于有灰条的图像的
# 我们需要对其进行修改，去除灰条的部分。 将box_xy、和box_wh调节成y_min,y_max,xmin,xmax
# 如果没有使用letterbox_image也需要将归一化后的box_xy, box_wh调整成相对于原图大小的
#------------------------------------------------------------------------------------------------------------#
boxes = yolo_correct_boxes(box_xy, box_wh, input_shape, image_shape, letterbox_image)
box_scores = box_confidence * box_class_probs
#-----------------------------------------------------------#
# 判断得分是否大于score_threshold
#-----------------------------------------------------------#
mask = box_scores >= confidence
max_boxes_tensor = K.constant(max_boxes, dtype='int32')
boxes_out = []
scores_out = []
classes_out = []
for c in range(num_classes):
#-----------------------------------------------------------#
# 取出所有box_scores >= score_threshold的框，和成绩
#-----------------------------------------------------------#
class_boxes = tf.boolean_mask(boxes, mask[:, c])
class_box_scores = tf.boolean_mask(box_scores[:, c], mask[:, c])
#-----------------------------------------------------------#
# 非极大抑制
# 保留一定区域内得分最大的框
#-----------------------------------------------------------#
nms_index = tf.image.non_max_suppression(class_boxes, class_box_scores, max_boxes_tensor, iou_threshold=nms_iou)
#-----------------------------------------------------------#
# 获取非极大抑制后的结果
# 下列三个分别是：框的位置，得分与种类
#-----------------------------------------------------------#
class_boxes = K.gather(class_boxes, nms_index)
class_box_scores = K.gather(class_box_scores, nms_index)
classes = K.ones_like(class_box_scores, 'int32') * c
boxes_out.append(class_boxes)
scores_out.append(class_box_scores)
classes_out.append(classes)
boxes_out = K.concatenate(boxes_out, axis=0)
scores_out = K.concatenate(scores_out, axis=0)
classes_out = K.concatenate(classes_out, axis=0)
return boxes_out, scores_out, classes_out
扫一扫
357
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
