yolov3之网络结构解析

参考网址:

https://blog.csdn.net/leviopku/article/details/82660381

yolov3之网络结构解析_第1张图片

yolov3之网络结构解析_第2张图片

YOLOV3采用了3个尺度的特征图(当输入为416×416时):13×13,26×26,52×52, yolov3每个位置使用3个先验框,所以使用k-means得到9个先验框,并将其划分到3个尺度特征图上,尺度更大的特征图使用更小的先验框。

网络基础结构

YOLOV3特征提取网络使用了残差模型,相比YOLOV2使用的Darknet-19,其包含53个卷积层,所以称为Darknet-53.

 

代码实现

 

Model = keras.models.Model
def tiny_yolo_body(inputs, num_anchors, num_classes):

    '''Create Tiny YOLO_v3 model CNN body in keras.'''
    x1 = compose(
            DarknetConv2D_BN_Leaky(16, (3,3)),
            MaxPooling2D(pool_size=(2,2), strides=(2,2), padding='same'),
            DarknetConv2D_BN_Leaky(32, (3,3)),
            MaxPooling2D(pool_size=(2,2), strides=(2,2), padding='same'),
            DarknetConv2D_BN_Leaky(64, (3,3)),
            MaxPooling2D(pool_size=(2,2), strides=(2,2), padding='same'),
            DarknetConv2D_BN_Leaky(128, (3,3)),
            MaxPooling2D(pool_size=(2,2), strides=(2,2), padding='same'),
            DarknetConv2D_BN_Leaky(256, (3,3)))(inputs)
    x2 = compose(
            MaxPooling2D(pool_size=(2,2), strides=(2,2), padding='same'),
            DarknetConv2D_BN_Leaky(512, (3,3)),
            MaxPooling2D(pool_size=(2,2), strides=(1,1), padding='same'),
            DarknetConv2D_BN_Leaky(1024, (3,3)),
            DarknetConv2D_BN_Leaky(256, (1,1)))(x1)
    y1 = compose(
            DarknetConv2D_BN_Leaky(512, (3,3)),
            DarknetConv2D(num_anchors*(num_classes+5), (1,1)))(x2)

    x2 = compose(
            DarknetConv2D_BN_Leaky(128, (1,1)),
            UpSampling2D(2))(x2)
    y2 = compose(
            Concatenate(),
            DarknetConv2D_BN_Leaky(256, (3,3)),
            DarknetConv2D(num_anchors*(num_classes+5), (1,1)))([x2,x1])

    return Model(inputs, [y1,y2])

DBL: 如图1左下角所示,也就是代码中的Darknetconv2d_BN_Leaky,是yolo_v3的基本组件。就是卷积+BN+Leaky relu。对于v3来说,BN和leaky relu已经是和卷积层不可分离的部分了(最后一层卷积除外),共同构成了最小组件。

代码实现

def DarknetConv2D_BN_Leaky(*args, **kwargs):
    """Darknet Convolution2D followed by BatchNormalization and LeakyReLU."""
    no_bias_kwargs = {'use_bias': False}
    no_bias_kwargs.update(kwargs)
    return compose(
        DarknetConv2D(*args, **no_bias_kwargs),
        BatchNormalization(),
        LeakyReLU(alpha=0.1))

resn:n代表数字,有res1,res2, … ,res8等等,表示这个res_block里含有多少个res_unit。这是yolo_v3的大组件,yolo_v3开始借鉴了ResNet的残差结构,使用这种结构可以让网络结构更深(从v2的darknet-19上升到v3的darknet-53,前者没有残差结构)。对于res_block的解释,可以在图1的右下角直观看到,其基本组件也是DBL。

代码实现

def resblock_body(x, num_filters, num_blocks):
    '''A series of resblocks starting with a downsampling Convolution2D'''
    # Darknet uses left and top padding instead of 'same' mode
    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 = compose(
                DarknetConv2D_BN_Leaky(num_filters//2, (1,1)),
                DarknetConv2D_BN_Leaky(num_filters, (3,3)))(x)
        x = Add()([x,y])
    return x

concat:张量拼接。将darknet中间层和后面的某一层的上采样进行拼接。拼接的操作和残差层add的操作是不一样的,拼接会扩充张量的维度,而add只是直接相加不会导致张量维度的改变。

你可能感兴趣的:(yolov3之网络结构解析)