第十一讲卷积神经网络（高级篇）

最新推荐文章于 2025-05-14 16:10:43 发布

爱编成不爱编程

最新推荐文章于 2025-05-14 16:10:43 发布

阅读量245

点赞数 1

分类专栏： Pytorch深度学习实践文章标签： cnn 人工智能神经网络

本文链接：https://blog.csdn.net/weixin_49933737/article/details/132136007

版权

Pytorch深度学习实践专栏收录该内容

12 篇文章

订阅专栏

第十一讲卷积神经网络（高级篇）学习与代码

B站刘二大人教学视频传送门：卷积神经网络（高级篇）

一、Inception Model

1、GoogLeNet（通过Pytorch实现）传送门GoogLeNet实现

其中有很多相同的结构叫Inception Model如下：

2、1X1卷积核（信息融合：即不同通道信息融合），作用为改变通道数：C1-->C2

3、改变通道后，如下图所示，很明显1X1卷积核改变通道后，可以减少运算数量。

4、Inception Model如下图所示：

5、做完四个分支卷积操作后，然后将他们拼起来，沿着C拼起来的代码为：

 outputs_=[branch1x1,branch5x5,branch3x3,branch_pool]
 # dim=1是因为(b,c,w,h),是要沿着c拼起来
 return torch.cat(outputs_,dim=1)

将这整个层抽象InceptionA

6、实现CNN-Inception-advance的代码

代码说明：1、将InceptionA层抽象为一个类方便直接调用

2、先进行两个卷积层，再进行两个Inception层，再进行一次池化层，最后线性

3、线性的大小1408 是通过view(batch,-1)之后张量计算出来的。

4、同样优化器使用的SGD，后进行训练测试

import torch
import matplotlib.pyplot as plt
from torchvision import transforms,datasets
from torch.utils.data import DataLoader
import torch.nn.functional as F
import os
os.environ["KMP_DUPLICATE_LIB_OK"]='True'

batch_size = 64
#归一化，均值和方差
transform = transforms.Compose([
    #Convert the PLI Image to Tensor
    transforms.ToTensor(),
    #参数为平均值和标准值
    transforms.Normalize((0.1307,),(0.3081,))
])

train_dataset = datasets.MNIST(root='../dataset/mnist/',
                             train=True,
                             download=True,
                             transform=transform
                             )
train_loader = DataLoader(train_dataset,
                        shuffle=True,
                        batch_size=batch_size
                        )
test_dataset = datasets.MNIST(root='../dataset/mnist/',
                             train=False,
                             download=True,
                             transform=transform)
test_loader = DataLoader(test_dataset,
                       shuffle=False,
                       batch_size=batch_size
                       )


class InceptionA(torch.nn.Module):
    def __init__(self,in_channels):
        super(InceptionA,self).__init__()
        self.branch1x1=torch.nn.Conv2d(in_channels,16,kernel_size=1)

        self.branch5x5_1=torch.nn.Conv2d(in_channels,16,kernel_size=1)
        self.branch5x5_2=torch.nn.Conv2d(16,24,kernel_size=5,padding=2)

        self.branchh3x3_1=torch.nn.Conv2d(in_channels,16, kernel_size=1)
        self.branchh3x3_2=torch.nn.Conv2d(16,24,kernel_size=3,padding=1)
        self.branchh3x3_3=torch.nn.Conv2d(24,24,kernel_size=3,padding=1)

        self.branch_pool=torch.nn.Conv2d(in_channels,24,kernel_size=1)

    def forward(self,x):
        branch1x1=self.branch1x1(x)

        branch5x5=self.branch5x5_1(x)
        branch5x5=self.branch5x5_2(branch5x5)

        branch3x3=self.branchh3x3_1(x)
        branch3x3=self.branchh3x3_2(branch3x3)
        branch3x3=self.branchh3x3_3(branch3x3)

        branch_pool=F.avg_pool2d(x,kernel_size=3,stride=1,padding=1)
        branch_pool= self.branch_pool(branch_pool)

        outputs_=[branch1x1,branch5x5,branch3x3,branch_pool]
        # dim=1是因为(b,c,w,h),是要沿着c拼起来
        return torch.cat(outputs_,dim=1)


class Net(torch.nn.Module):
    def __init__(self):
        super(Net,self).__init__()
        self.conv1=torch.nn.Conv2d(1,10,kernel_size=5)
        self.conv2=torch.nn.Conv2d(88,20,kernel_size=5)

        self.incep1=InceptionA(in_channels=10)
        self.incep2=InceptionA(in_channels=20)

        self.mp=torch.nn.MaxPool2d(2)
        self.fc=torch.nn.Linear(1408,10)

    def forward(self,x):
        in_size=x.size(0)
        x=F.relu(self.mp(self.conv1(x)))
        x=self.incep1(x)
        x=F.relu(self.mp(self.conv2(x)))
        x=self.incep2(x)
        x=x.view(in_size,-1)
        x=self.fc(x)
        return x

model=Net()

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
model.to(device)

criterion=torch.nn.CrossEntropyLoss()
optimizer=torch.optim.SGD(model.parameters(),lr=0.01,momentum=0.5)


def train(epoch):
    running_loss = 0.0
    for batch_idx, data in enumerate(train_loader, 0):
        inputs, target = data

        #Send the inputs and targets at every step to the GPU.
        inputs, target = inputs.to(device), target.to(device)

        optimizer.zero_grad()
        # forward + backward + update
        outputs = model(inputs)
        loss = criterion(outputs, target)
        loss.backward()
        optimizer.step()
        running_loss += loss.item()
        if batch_idx % 300 == 299:
            print('[%d, %5d] loss: %.3f' % (epoch + 1, batch_idx + 1, running_loss / 2000))
            running_loss = 0.

epoch_list=[]
accept_list=[]
def test2():
    correct = 0
    total = 0
    with torch.no_grad():
        for data in test_loader:
            images, target = data

            #Send the inputs and targets at every step to the GPU.
            images, target = images.to(device), target.to(device)

            outputs = model(images)
            _, predicted = torch.max(outputs.data, dim=1)
            total += target.size(0)
            correct += (predicted == target).sum().item()
    accept_list.append(100 * correct / total)
    print('Accuracy on test set: %d %% [%d/%d]' % (100 * correct / total, correct, total))
if __name__=='__main__':
    for epoch in range(10):
        train(epoch)
        test2()
        epoch_list.append(epoch)
plt.plot(epoch_list,accept_list)
plt.xlabel("epoch")
plt.rcParams['font.sans-serif'] = ['SimHei'] # 设置中文字体
plt.ylabel("成功率 单位为 %")
plt.show()

运行效果图：

二、Residual net

视频截图分析：

1、Residual net 是为了解决一直3X3卷积造成梯度消失的问题。

2、跳连接，H(x) = F(x) + x,张量维度必须一样，加完后再激活。

（导数+x再求导就不会只有导数，无限叠下去，那么肯定不会造成梯度消失）

为什么可以消除：因为求导过后后面会+1，若倒数非常小的时候则整个值再1附近，那么若干相乘，就能非常好的去解决梯度消失问题。如图右边所示：

3、ResidualBlock（）

4、源码

整个过程：

看过前面的基础篇，应该耳熟能详，毕竟更换模型的作业也做过：

代码说明：1、与Incepiton Model 一样形成一个ResidualBlock Model 封装

2、过程大概时一次卷积，一次Residual ，再进行一次卷积，再进行Residual最后线性

3、如果只用CPU，将注释有Send开头的删掉，他的主要租用是将数据放入GPU。

代码实现：

import torch
import torch.nn.functional as F
from torchvision import transforms,datasets
from torch.utils.data import DataLoader
import matplotlib.pyplot as plt

import os
os.environ["KMP_DUPLICATE_LIB_OK"]='True'

batch_size = 64
transform = transforms.Compose([
    #Convert the PLI Image to Tensor
    transforms.ToTensor(),
    #参数为平均值和标准值
    transforms.Normalize((0.1307,),(0.3081,))
])

train_dataset = datasets.MNIST(root='../dataset/mnist/',
                             train=True,
                             download=True,
                             transform=transform
                             )
train_loader = DataLoader(train_dataset,
                        shuffle=True,
                        batch_size=batch_size
                        )
test_dataset = datasets.MNIST(root='../dataset/mnist/',
                             train=False,
                             download=True,
                             transform=transform)
test_loader = DataLoader(test_dataset,
                       shuffle=False,
                       batch_size=batch_size
                         )
class ResidualBlock(torch.nn.Module):
    def __init__(self,channels):
        super(ResidualBlock,self).__init__()
        self.channels=channels
        self.conv1=torch.nn.Conv2d(channels,channels,kernel_size=3,padding=1)

        self.conv2=torch.nn.Conv2d(channels,channels,kernel_size=3,padding=1)


    def forward(self,x):
        y=F.relu(self.conv1(x))
        y=self.conv2(y)
        return F.relu(x+y)

class Net(torch.nn.Module):
    def __init__(self):
        super(Net,self).__init__()
        self.conv1=torch.nn.Conv2d(1,16,kernel_size=5)
        self.conv2=torch.nn.Conv2d(16,32,kernel_size=5)
        self.mp=torch.nn.MaxPool2d(2)

        self.rblock1=ResidualBlock(16)
        self.rblock2=ResidualBlock(32)

        self.fc=torch.nn.Linear(512,10)
    def forward(self,x):
        in_size=x.size(0)
        x=self.mp(F.relu(self.conv1(x)))
        x=self.rblock1(x)
        x=self.mp(F.relu(self.conv2(x)))
        x=self.rblock2(x)
        x=x.view(in_size,-1)
        x=self.fc(x)
        return x

model = Net()

# #Convert parameters and buffers of all modules to CUDA Tensor.
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
model.to(device)


#多分类CrossEntropyLoss  二分类则用BCELoss
criterion=torch.nn.CrossEntropyLoss()
optimizer=torch.optim.SGD(model.parameters(),lr=0.01,momentum=0.5)


def train(epoch):
    running_loss = 0.0
    for batch_idx, data in enumerate(train_loader, 0):
        inputs, target = data

        #Send the inputs and targets at every step to the GPU.
        inputs, target = inputs.to(device), target.to(device)

        optimizer.zero_grad()
        # forward + backward + update
        outputs = model(inputs)
        loss = criterion(outputs, target)
        loss.backward()
        optimizer.step()
        running_loss += loss.item()
        if batch_idx % 300 == 299:
            print('[%d, %5d] loss: %.3f' % (epoch + 1, batch_idx + 1, running_loss / 2000))
            running_loss = 0.

epoch_list=[]
accept_list=[]
def test2():
    correct = 0
    total = 0
    with torch.no_grad():
        for data in test_loader:
            images, target = data

            #Send the inputs and targets at every step to the GPU.
            images, target = images.to(device), target.to(device)

            outputs = model(images)
            _, predicted = torch.max(outputs.data, dim=1)
            total += target.size(0)
            correct += (predicted == target).sum().item()
    accept_list.append(100 * correct / total)
    print('Accuracy on test set: %d %% [%d/%d]' % (100 * correct / total, correct, total))
if __name__=='__main__':
    for epoch in range(10):
        train(epoch)
        test2()
        epoch_list.append(epoch)
plt.plot(epoch_list,accept_list)
plt.xlabel("epoch")
plt.rcParams['font.sans-serif'] = ['SimHei'] # 设置中文字体
plt.ylabel("成功率 单位为 %")
plt.show()

运行结果：

.......

Accuracy on test set: 98 % [9866/10000]
[7, 300] loss: 0.005
[7, 600] loss: 0.005
[7, 900] loss: 0.005
Accuracy on test set: 98 % [9881/10000]
[8, 300] loss: 0.005
[8, 600] loss: 0.004
[8, 900] loss: 0.004
Accuracy on test set: 98 % [9889/10000]
[9, 300] loss: 0.004
[9, 600] loss: 0.004
[9, 900] loss: 0.004
Accuracy on test set: 98 % [9883/10000]
[10, 300] loss: 0.003
[10, 600] loss: 0.003
[10, 900] loss: 0.004
Accuracy on test set: 98 % [9864/10000]

进程已结束，退出代码为 0

运行效果：