cvpr2017 Densenet

正在做实验，要用到特征图连接，不知道怎么下手，以及是否符合反向传播，所以过来重新看下densenet以及它的代码。
下面论文内容主要引用博客https://blog.csdn.net/u014380165/article/details/75142664

densenet
上面是主要模型图
在传统的卷积神经网络中，如果你有L层，那么就会有L个连接，但是在DenseNet中，会有L(L+1)/2个连接。简单讲，就是每一层的输入来自前面所有层的输出。如下图：x0是input，H1的输入是x0（input），H2的输入是x0和x1（x1是H1的输出）
公式

然后呢 densenet实际工作就是把之前的特征图进行融合在了一起，作为新的输入。

图片.png

然后我们来看下代码，在github找了个新一点的代码https://github.com/andreasveit/densenet-pytorch
模型代码只有100多行过一点，就全看一下好了。

class BasicBlock(nn.Module):
    def __init__(self, in_planes, out_planes, dropRate=0.0):
        super(BasicBlock, self).__init__()
        self.bn1 = nn.BatchNorm2d(in_planes)
        self.relu = nn.ReLU(inplace=True)
        self.conv1 = nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=1,
                               padding=1, bias=False)
        self.droprate = dropRate
    def forward(self, x):
        out = self.conv1(self.relu(self.bn1(x)))
        if self.droprate > 0:
            out = F.dropout(out, p=self.droprate, training=self.training)
return torch.cat([x, out], 1)

BasicBlock： bn,relu,conv三件套再加个可选择的dropout，然后用torch.cat在通道上进行融合。

class BottleneckBlock(nn.Module):
    def __init__(self, in_planes, out_planes, dropRate=0.0):
        super(BottleneckBlock, self).__init__()
        inter_planes = out_planes * 4
        self.bn1 = nn.BatchNorm2d(in_planes)
        self.relu = nn.ReLU(inplace=True)
        self.conv1 = nn.Conv2d(in_planes, inter_planes, kernel_size=1, stride=1,
                               padding=0, bias=False)
        self.bn2 = nn.BatchNorm2d(inter_planes)
        self.conv2 = nn.Conv2d(inter_planes, out_planes, kernel_size=3, stride=1,
                               padding=1, bias=False)
        self.droprate = dropRate
    def forward(self, x):
        out = self.conv1(self.relu(self.bn1(x)))
        if self.droprate > 0:
            out = F.dropout(out, p=self.droprate, inplace=False, training=self.training)
        out = self.conv2(self.relu(self.bn2(out)))
        if self.droprate > 0:
            out = F.dropout(out, p=self.droprate, inplace=False, training=self.training)
        return torch.cat([x, out], 1)

BottleneckBlock似乎就是把BasicBlock中间三件套重复了2遍？？？

class TransitionBlock(nn.Module):
    def __init__(self, in_planes, out_planes, dropRate=0.0):
        super(TransitionBlock, self).__init__()
        self.bn1 = nn.BatchNorm2d(in_planes)
        self.relu = nn.ReLU(inplace=True)
        self.conv1 = nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=1,
                               padding=0, bias=False)
        self.droprate = dropRate
    def forward(self, x):
        out = self.conv1(self.relu(self.bn1(x)))
        if self.droprate > 0:
            out = F.dropout(out, p=self.droprate, inplace=False, training=self.training)
return F.avg_pool2d(out, 2)

TransitionBlock在3件套加了个AvgPool2d() ，二维平均池化层，用来减小featuremap数量以及pooling提取出图片本来的特征

class DenseBlock(nn.Module):
    def __init__(self, nb_layers, in_planes, growth_rate, block, dropRate=0.0):
        super(DenseBlock, self).__init__()
        self.layer = self._make_layer(block, in_planes, growth_rate, nb_layers, dropRate)
    def _make_layer(self, block, in_planes, growth_rate, nb_layers, dropRate):
        layers = []
        for i in range(nb_layers):
            layers.append(block(in_planes+i*growth_rate, growth_rate, dropRate))
        return nn.Sequential(*layers)
    def forward(self, x):
      return self.layer(x)

DenseBlock定义了denset-net连接方式，通用的layers层

class DenseNet3(nn.Module):
    def __init__(self, depth, num_classes, growth_rate=12,
                 reduction=0.5, bottleneck=True, dropRate=0.0):
        super(DenseNet3, self).__init__()
        in_planes = 2 * growth_rate
        n = (depth - 4) / 3
        if bottleneck == True:
            n = n/2
            block = BottleneckBlock
        else:
            block = BasicBlock
        n = int(n)
        # 1st conv before any dense block
        self.conv1 = nn.Conv2d(3, in_planes, kernel_size=3, stride=1,
                               padding=1, bias=False)
        # 1st block
        self.block1 = DenseBlock(n, in_planes, growth_rate, block, dropRate)
        in_planes = int(in_planes+n*growth_rate)
        self.trans1 = TransitionBlock(in_planes, int(math.floor(in_planes*reduction)), dropRate=dropRate)
        in_planes = int(math.floor(in_planes*reduction))
        # 2nd block
        self.block2 = DenseBlock(n, in_planes, growth_rate, block, dropRate)
        in_planes = int(in_planes+n*growth_rate)
        self.trans2 = TransitionBlock(in_planes, int(math.floor(in_planes*reduction)), dropRate=dropRate)
        in_planes = int(math.floor(in_planes*reduction))
        # 3rd block
        self.block3 = DenseBlock(n, in_planes, growth_rate, block, dropRate)
        in_planes = int(in_planes+n*growth_rate)
        # global average pooling and classifier
        self.bn1 = nn.BatchNorm2d(in_planes)
        self.relu = nn.ReLU(inplace=True)
        self.fc = nn.Linear(in_planes, num_classes)
        self.in_planes = in_planes

        for m in self.modules():
            if isinstance(m, nn.Conv2d):
                n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
                m.weight.data.normal_(0, math.sqrt(2. / n))
            elif isinstance(m, nn.BatchNorm2d):
                m.weight.data.fill_(1)
                m.bias.data.zero_()
            elif isinstance(m, nn.Linear):
                m.bias.data.zero_()
    def forward(self, x):
        out = self.conv1(x)
        out = self.trans1(self.block1(out))
        out = self.trans2(self.block2(out))
        out = self.block3(out)
        out = self.relu(self.bn1(out))
        out = F.avg_pool2d(out, 8)
        out = out.view(-1, self.in_planes)
        return self.fc(out)

bottleBlock和basicblock二选一，作为重复的最小的快
1st conv 提取特征，把feature map数量扩大
self.conv1 = nn.Conv2d(3, in_planes, kernel_size=3, stride=1,
padding=1, bias=False)
若干个(dense block和TransitionBlock)+全连接