ecv 2018: receptivefieldblocknetforaccurateandfastobjectdetection提出了一种新的特征提取模块——RFB，本文的出发点是模拟人的视觉感受野，增强网络特征提取能力
RFB效果的图像如图所示。 中央虚线的框部分为RFB结构。 RFB结构主要有两个特点。
1、由不同大小卷积核卷积层组成的多分枝结构。 这一部分可以参考Inception结构。 在Figure2的RFB结构中，也用尺寸不同的圆形表示了尺寸不同的卷积核的卷积层。
2、引入了dilated卷积层。 dilated卷积层以前被应用于分割算法深度lab，主要作用也是增加感受区，与deformable卷积有区别。
在RFB结构中，dilated卷积层的参数用不同的rate表示。 在结构中最后连接不同大小和rate卷积层的输出，达到融合不同特征的目的。 在结构中重叠展示三种不同大小和颜色的输出。 最后一列将融合的特征与人类的视觉感觉野进行了对比。 从图中可以看出，非常近。 这也是这篇文章的出发点，换句话说，就是模拟人类视觉的感觉区域进行RFB结构的设计。
下图为2种RFB结构的示意图。 (a )为射频b，总体结构借鉴了Inception的思想，主要区别在于引入3个延迟卷积层(如33conv，rate=1)，也是本文增加感受区的主要方式之一(b )为射频RFB-s。 RFB-s与RFB相比主要有两个改进，一方面用33卷积层代替55卷积层，另一方面用13和31卷积层代替33卷积层。 主要目的应该是减少计算量，类似于改进Inception后期版本的Inception结构。
RFB代码的实现
classbasicconv(nn.module ) :def_init_(self，in_planes，out_planes，kernel_size，stride=1，bias kernel_size=kernel_size，stride=stride，padding=padding，dilation=dilation，groups=groups，bias=bias ) affine=True ) ifbnelsenoneself.relu=nn.relu (in place=true ) ifreluelsenonedefforward ) self， x65:x=self.conv(x ) if self.bnisnotnone : x=self.bn ) x ) if self.reluisnotnone 3360 x=self.relu (x ) rell in_planes，out_planes，stride=1，scale=0.1，visual=1) 3360super (基本反射器)。 self(_init_ ) self.scale=scale self.out _ channels=out _ planes inter _ planes=in _ planes//8see lf2* ines basicconv(2*inter_planes，2*inter_planes，Kernel relu=False ) (self.branch1=nn.sequential (basic conv )
, BasicConv(2*inter_planes, 2*inter_planes, kernel_size=3, stride=1, padding=visual+1, dilation=visual+1, relu=False) ) self.branch2 = nn.Sequential( BasicConv(in_planes, inter_planes, kernel_size=1, stride=1), BasicConv(inter_planes, (inter_planes//2)*3, kernel_size=3, stride=1, padding=1), BasicConv((inter_planes//2)*3, 2*inter_planes, kernel_size=3, stride=stride, padding=1), BasicConv(2*inter_planes, 2*inter_planes, kernel_size=3, stride=1, padding=2*visual+1, dilation=2*visual+1, relu=False) ) self.ConvLinear = BasicConv(6*inter_planes, out_planes, kernel_size=1, stride=1, relu=False) self.shortcut = BasicConv(in_planes, out_planes, kernel_size=1, stride=stride, relu=False) self.relu = nn.ReLU(inplace=False) def forward(self,x): x0 = self.branch0(x) x1 = self.branch1(x) x2 = self.branch2(x) out = torch.cat((x0,x1,x2),1) out = self.ConvLinear(out) short = self.shortcut(x) out = out*self.scale + short out = self.relu(out) return outclass BasicRFB_a(nn.Module): def __init__(self, in_planes, out_planes, stride=1, scale = 0.1): super(BasicRFB_a, self).__init__() self.scale = scale self.out_channels = out_planes inter_planes = in_planes //4 self.branch0 = nn.Sequential( BasicConv(in_planes, inter_planes, kernel_size=1, stride=1), BasicConv(inter_planes, inter_planes, kernel_size=3, stride=1, padding=1,relu=False) ) self.branch1 = nn.Sequential( BasicConv(in_planes, inter_planes, kernel_size=1, stride=1), BasicConv(inter_planes, inter_planes, kernel_size=(3,1), stride=1, padding=(1,0)), BasicConv(inter_planes, inter_planes, kernel_size=3, stride=1, padding=3, dilation=3, relu=False) ) self.branch2 = nn.Sequential( BasicConv(in_planes, inter_planes, kernel_size=1, stride=1), BasicConv(inter_planes, inter_planes, kernel_size=(1,3), stride=stride, padding=(0,1)), BasicConv(inter_planes, inter_planes, kernel_size=3, stride=1, padding=3, dilation=3, relu=False) ) self.branch3 = nn.Sequential( BasicConv(in_planes, inter_planes//2, kernel_size=1, stride=1), BasicConv(inter_planes//2, (inter_planes//4)*3, kernel_size=(1,3), stride=1, padding=(0,1)), BasicConv((inter_planes//4)*3, inter_planes, kernel_size=(3,1), stride=stride, padding=(1,0)), BasicConv(inter_planes, inter_planes, kernel_size=3, stride=1, padding=5, dilation=5, relu=False) ) self.ConvLinear = BasicConv(4*inter_planes, out_planes, kernel_size=1, stride=1, relu=False) self.shortcut = BasicConv(in_planes, out_planes, kernel_size=1, stride=stride, relu=False) self.relu = nn.ReLU(inplace=False) def forward(self,x): x0 = self.branch0(x) x1 = self.branch1(x) x2 = self.branch2(x) x3 = self.branch3(x) out = torch.cat((x0,x1,x2,x3),1) out = self.ConvLinear(out) short = self.shortcut(x) out = out*self.scale + short out = self.relu(out) return out