【从零学习PyTorch】 如何残差网络resnet作为pre-model +代码讲解+残差网络resnet是个啥

看的多个Kaggle上 图片分类比赛 的代码,发现基本都会选择resnet网络作为前置网络进行训练,那么如何实现这个呢?
本文主要分为两个部分

  • 第一个部分讲解如何使用PyTorch来实现前置网络的设置,以及参数的下载和导入
  • 第二个部分简单讲一下resnet运行的原理。

第一部分:实现

有一个非常好用的库,叫做torchvision。
这个是torchvision的官方文档
这个库有三个部分:

  • torchvision.datasets
  • torchvision.models
  • torchvision.transforms
import torchvision
model = torchvision.models.resnet50(pretrained=False)

这个model,就跟我之前讲的自己写一个class Net(nn.module),之后net = Net()的net是一样的!经典的网络一般是不需要重新写的,我们需要站在他们的肩膀上,简单修改就可以了。
从零学习pytorch 第5课 PyTorch模型搭建三要素

代码中,pretrained=False是不加载参数,也就是说这个model中的参数是初始化的而不是训练好的。这里我们自然是需要之前训练的参数作为我们的初始参数。所以我们改成True即可,这里会自动下载一个参数文件。

import torchvision
model = torchvision.models.resnet50(pretrained=True)

这里可以看一下这个的源码,resnet18,resnet50,resnet101其实模型构建都一样,就是参数略有差别。如果pretrained是True的时候,会自动调用model.load_state_dict()这个函数,其实就是加载模型的函数。模型文件会从一个model_zoo.load_url下载参数。

def resnet18(pretrained=False, **kwargs):
    """Constructs a ResNet-18 model.
    Args:
        pretrained (bool): If True, returns a model pre-trained on ImageNet
    """
    model = ResNet(BasicBlock, [2, 2, 2, 2], **kwargs)
    if pretrained:
        model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))
    return model
def resnet50(pretrained=False, **kwargs):
    """Constructs a ResNet-50 model.
    Args:
        pretrained (bool): If True, returns a model pre-trained on ImageNet
    """
    model = ResNet(Bottleneck, [3, 4, 6, 3], **kwargs)
    if pretrained:
        model.load_state_dict(model_zoo.load_url(model_urls['resnet50']))
    return model
model_urls = {
    'resnet18': 'https://download.pytorch.org/models/resnet18-5c106cde.pth',
    'resnet34': 'https://download.pytorch.org/models/resnet34-333f7ec4.pth',
    'resnet50': 'https://download.pytorch.org/models/resnet50-19c8e357.pth',
    'resnet101': 'https://download.pytorch.org/models/resnet101-5d3b4d8f.pth',
    'resnet152': 'https://download.pytorch.org/models/resnet152-b121ed2d.pth',
}

我们可以看到,resnet18啊50啊101啊10086啊都是修改了ResNet的一个参数,那这个参数什么意思呢?
我们更进一步,看看ResNet的代码:

class ResNet(nn.Module):
    def __init__(self, block, layers, num_classes=1000):
        self.inplanes = 64
        super(ResNet, self).__init__()
        self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3,
                               bias=False)
        self.bn1 = nn.BatchNorm2d(64)
        self.relu = nn.ReLU(inplace=True)
        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
        self.layer1 = self._make_layer(block, 64, layers[0])
        self.layer2 = self._make_layer(block, 128, layers[1], stride=2)
        self.layer3 = self._make_layer(block, 256, layers[2], stride=2)
        self.layer4 = self._make_layer(block, 512, layers[3], stride=2)
        self.avgpool = nn.AvgPool2d(7, stride=1)
        self.fc = nn.Linear(512 * block.expansion, num_classes)
        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)) # 卷积参数变量初始化
            elif isinstance(m, nn.BatchNorm2d):
                m.weight.data.fill_(1) # BN参数初始化
                m.bias.data.zero_()
    def _make_layer(self, block, planes, blocks, stride=1):
        downsample = None
        if stride != 1 or self.inplanes != planes * block.expansion:
            downsample = nn.Sequential(
                nn.Conv2d(self.inplanes, planes * block.expansion,
                          kernel_size=1, stride=stride, bias=False),
                nn.BatchNorm2d(planes * block.expansion),
            )
        layers = []
        layers.append(block(self.inplanes, planes, stride, downsample))
        self.inplanes = planes * block.expansion
        for i in range(1, blocks):
            layers.append(block(self.inplanes, planes))
        return nn.Sequential(*layers)
    def forward(self, x):
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)
        x = self.maxpool(x)
        x = self.layer1(x)
        x = self.layer2(x)
        x = self.layer3(x)
        x = self.layer4(x)
        x = self.avgpool(x)
        x = x.view(x.size(0), -1)
        x = self.fc(x)
        return x

我们来分析一下这段代码:

def __init__(self, block, layers, num_classes=1000):

在初始化的时候,有两个参数,block和layers。

  • block有两种,一种是Bottleneck,一种是Basicblock。在resnet18和resnet34中调用的是BasicBlock类,在resnet50,resnet101,resnet152中调用的是Bottleneck类。
  • layers是包含四个元素的列表,每个元素分别是_make_layer生成四个大的layer中包含的resdual子结构的个数。resnet50中式[3,4,6,3]
def _make_layer(self, block, planes, blocks, stride=1):

_make_layer包含四个参数,第一个就是block类型,第二个是planes也就是输出的channel的数量,第三个blocks是每个blcoks中包含多少的子结构,其实就是layers所存储的数组,第四个是步长

接下来我们看一下两种block的源码

class Bottleneck(nn.Module):
    expansion = 4
    def __init__(self, inplanes, planes, stride=1, downsample=None):
        super(Bottleneck, self).__init__()
        self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, bias=False)
        self.bn1 = nn.BatchNorm2d(planes)
        self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=stride,
                               padding=1, bias=False)
        self.bn2 = nn.BatchNorm2d(planes)
        self.conv3 = nn.Conv2d(planes, planes * 4, kernel_size=1, bias=False)
        self.bn3 = nn.BatchNorm2d(planes * 4)
        self.relu = nn.ReLU(inplace=True)
        self.downsample = downsample
        self.stride = stride
    def forward(self, x):
        residual = x
        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)
        out = self.conv2(out)
        out = self.bn2(out)
        out = self.relu(out)
        out = self.conv3(out)
        out = self.bn3(out)
        if self.downsample is not None:
            residual = self.downsample(x)
        out += residual
        out = self.relu(out)
        return out

两种block:Bottleneck和Basicblock两个block都继承了torch.nn.Module。(这里总算是回到了torch模型三定义里面了,复杂的模型果然不一样)
简单看了一下,就是:
Conv2d->BatchNorm2d->ReLU->Conv2d->BatchNorm2d->ReLU->Conv2d->BatchNorm2d(->downsample)->ReLU这样的。之后画个图记忆深刻一点,虽然不知道记这些东西有没有用吧,跟同专业的人装装逼应该没问题。

【从零学习PyTorch】 如何残差网络resnet作为pre-model +代码讲解+残差网络resnet是个啥

def conv3x3(in_planes, out_planes, stride=1):
    """3x3 convolution with padding"""
    return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
                     padding=1, bias=False)
class BasicBlock(nn.Module):
    expansion = 1
    def __init__(self, inplanes, planes, stride=1, downsample=None):
        super(BasicBlock, self).__init__()
        self.conv1 = conv3x3(inplanes, planes, stride)
        self.bn1 = nn.BatchNorm2d(planes)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = conv3x3(planes, planes)
        self.bn2 = nn.BatchNorm2d(planes)
        self.downsample = downsample
        self.stride = stride
    def forward(self, x):
        residual = x
        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)
        out = self.conv2(out)
        out = self.bn2(out)
        if self.downsample is not None:
            residual = self.downsample(x)
        out += residual
        out = self.relu(out)
        return out

【从零学习PyTorch】 如何残差网络resnet作为pre-model +代码讲解+残差网络resnet是个啥
Basicblock少了一个卷积层,相比Bottleneck。

resnet18模型流程

resnet18调用的参数是model = ResNet(BasicBlock,[2,2,2,2],**kwargs)
在代码中,最难理解的就是self.layer1 = self.._make_layer(block,64,layers[0])
因为是resnet18,所以block是Basicblock,layer[0]=2

核心就是:layers.append(block(self.inplanes,planes, stride, downsample))
sefl.inplanes = 64,planes = 64,strid=1
然后就按照上面那个流程图,构造出一个两层卷积的Basicblock。

for i in range(1, blocks):
            layers.append(block(self.inplanes, planes))

然后这个blocks就是传入的参数layers[0],
总之,layers中有layer[0]个blocks,(layer[0]不能是0.)

我们再看layer2中,传入的参数是stride=2,这倒是downsample不再是None了,然后发现:
每一层中,如果stride=1,就没有downsamle,如果stride=2,就有downsample,而且是这个layers中第一个block才有downsample,之后的block依然没有downsample=

从layer2到layer4,没一层都会增加一本的channel,所以residual,也就是block会采用下采样。而对于每一层而言,channel都是相同的,因为在每一层的第一个block的第一个卷积层,就会把channel放大一倍,然后这一层的之后都是多少channel进去,多少channel出来。

总结

【从零学习PyTorch】 如何残差网络resnet作为pre-model +代码讲解+残差网络resnet是个啥
上图是一个Resnet34的结构图

现在我算是彻底明白34到底是怎么计算来的

模型一开始有一个卷积层,最后有一个全连接层,这是2层

然后【3,4,6,3】是Resnet34的参数,使用Basicblock,每个block2个卷积层

所以(3+4+6+3)*2+2 = 34

对于Resnet50,其实输入参数一样,还是【3,4,6,3】,只是Resnet50使用的Bottleneck,是三层卷积层,所以是(3+4+6+3)*3+2 = 50

上一篇:经典网络结构(六)ResNet


下一篇:ResNet小结