ResNet

網(wǎng)絡(luò)結(jié)構(gòu)

代碼

#------------------------------用50行代碼搭建ResNet------------------------------------------- from torch import nn import torch as t from torch.nn import functional as Fclass ResidualBlock(nn.Module):#實(shí)現(xiàn)子module: Residual Blockdef __init__(self,inchannel,outchannel,stride=1,shortcut=None):super(ResidualBlock,self).__init__()self.left=nn.Sequential(nn.Conv2d(inchannel,outchannel,3,stride,1,bias=False),nn.BatchNorm2d(outchannel),nn.ReLU(inplace=True),nn.Conv2d(outchannel,outchannel,3,1,1,bias=False),nn.BatchNorm2d(outchannel))self.right=shortcutdef forward(self,x):out=self.left(x)residual=x if self.right is None else self.right(x)out+=residualreturn F.relu(out)class ResNet(nn.Module):#實(shí)現(xiàn)主module:ResNet34#ResNet34包含多個(gè)layer,每個(gè)layer又包含多個(gè)residual block#用子module實(shí)現(xiàn)residual block , 用 _make_layer 函數(shù)實(shí)現(xiàn)layerdef __init__(self,num_classes=1000):super(ResNet,self).__init__()self.pre=nn.Sequential(nn.Conv2d(3,64,7,2,3,bias=False),nn.BatchNorm2d(64),nn.ReLU(inplace=True),nn.MaxPool2d(3,2,1))#重復(fù)的layer,分別有3,4,6,3個(gè)residual blockself.layer1=self._make_layer(64,64,3)self.layer2=self._make_layer(64,128,4,stride=2)self.layer3=self._make_layer(128,256,6,stride=2)self.layer4=self._make_layer(256,512,3,stride=2)#分類用的全連接self.fc=nn.Linear(512,num_classes)def _make_layer(self,inchannel,outchannel,block_num,stride=1):#構(gòu)建layer,包含多個(gè)residual blockshortcut=nn.Sequential(nn.Conv2d(inchannel,outchannel,1,stride,bias=False),nn.BatchNorm2d(outchannel))layers=[ ]layers.append(ResidualBlock(inchannel,outchannel,stride,shortcut))for i in range(1,block_num):layers.append(ResidualBlock(outchannel,outchannel))return nn.Sequential(*layers)def forward(self,x):x=self.pre(x)x=self.layer1(x)x=self.layer2(x)x=self.layer3(x)x=self.layer4(x)x=F.avg_pool2d(x,7)x=x.view(x.size(0),-1)return self.fc(x)resnet = ResNet() print(resnet)

輸出結(jié)果如下：

ResNet((pre): Sequential((0): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)(1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False))(layer1): Sequential((0): ResidualBlock((left): Sequential((0): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True))(right): Sequential((0): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)(1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(1): ResidualBlock((left): Sequential((0): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(2): ResidualBlock((left): Sequential((0): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True))))(layer2): Sequential((0): ResidualBlock((left): Sequential((0): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)(1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True))(right): Sequential((0): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2), bias=False)(1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(1): ResidualBlock((left): Sequential((0): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(2): ResidualBlock((left): Sequential((0): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(3): ResidualBlock((left): Sequential((0): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True))))(layer3): Sequential((0): ResidualBlock((left): Sequential((0): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)(1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True))(right): Sequential((0): Conv2d(128, 256, kernel_size=(1, 1), stride=(2, 2), bias=False)(1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(1): ResidualBlock((left): Sequential((0): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(2): ResidualBlock((left): Sequential((0): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(3): ResidualBlock((left): Sequential((0): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(4): ResidualBlock((left): Sequential((0): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(5): ResidualBlock((left): Sequential((0): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True))))(layer4): Sequential((0): ResidualBlock((left): Sequential((0): Conv2d(256, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)(1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True))(right): Sequential((0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)(1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(1): ResidualBlock((left): Sequential((0): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)))(2): ResidualBlock((left): Sequential((0): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)(2): ReLU(inplace=True)(3): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)(4): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True))))(fc): Linear(in_features=512, out_features=1000, bias=True) )

論文一些注意點(diǎn)

擬合高維

深度神經(jīng)網(wǎng)絡(luò)可以擬合高維非線性函數(shù)

• 可以擬合 $H(x)$，也可以擬合 $H(x)?x$。
• 令 $F(x)=H(x)?x$，則 $H(x)=F(x)+x$。
• 網(wǎng)絡(luò)學(xué)習(xí)的難易程度不同。
• 解決梯度消失問題，增加的網(wǎng)絡(luò)層學(xué)習(xí)成恒等映射，則不會(huì)提升訓(xùn)練誤差。

跳層的恒等映射

殘差塊
$$y=F(x,\{w_i\})+x$$
其中，$F$：需要學(xué)習(xí)的殘差映射，維度與 $x$ 一致；$F+x$：跳躍連接，逐一加和，最后二者的輸出經(jīng)過函數(shù) ReLU；

沒有額外參數(shù)，不增加復(fù)雜度。
$F$ 包含兩個(gè)或兩個(gè)以上網(wǎng)絡(luò)層，否則表現(xiàn)為線性層 $y=w_{1}x+x$。

殘差網(wǎng)絡(luò)

兩個(gè)簡單的設(shè)計(jì)原則：

具有相同輸出特征圖大小的網(wǎng)絡(luò)卷積層卷積核數(shù)量相同；

特征圖大小減半，則卷積數(shù)量增加一倍，保證時(shí)間復(fù)雜度。

基于普通網(wǎng)絡(luò)，插入跳躍連接。
維度增加時(shí)，增加零輸出，或 1x1 卷積改變維度。

Ref

Pytorch學(xué)習(xí)（三）–用50行代碼搭建ResNet

Pytorch實(shí)戰(zhàn)2：ResNet-18實(shí)現(xiàn)Cifar-10圖像分類（測(cè)試集分類準(zhǔn)確率95.170%）

總結(jié)

以上是生活随笔為你收集整理的PyTorch 实现经典模型5：ResNet的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。