【PyTorch深度学习实践】09_卷积神经网络基础

1.卷积操作

卷积神经网络概览

1.1 卷积操作

输入通道数=卷积核通道数，卷积核个数=输出通道数

1.2 padding-填充

padding是为了让源图像最外一圈或多圈像素（取决于kernel的尺寸），能够被卷积核中心取到。
这里有个描述很重要：想要使源图像（1,1）的位置作为第一个与kernel中心重合，参与计算的像素，想想看padding需要扩充多少层，这样就很好计算了

1.3 stride-步长

stride操作指的是每次kernel窗口滑动的步长，默认值是1

1.4 pooling-池化

以最大池化为例

卷积神经网络示例

1.5 基础版CNN代码示例

import torch
in_channels, out_channels=5, 10
width, height = 100, 100    # 图像大小
kernel_size = 3     # 卷积核大小
batch_size = 1  # 所有输入pytorch中的data必须是小批量

input = torch.randn(
    batch_size,  # B 表明是小批量第几个
    in_channels,  # n 输入通道数
    width,      # W 宽
    height      # H 高
)

conv_layer = torch.nn.Conv2d(
    in_channels,    # 输入通道数
    out_channels,   # 输出通道数
    kernel_size=kernel_size  # 内核大小
)

output = conv_layer(input)

print(input.shape)
print(output.shape)
print(conv_layer.weight.shape)

1.6 完整CNN代码示例

import matplotlib.pyplot as plt
import torch

from torchvision import datasets, transforms
from torch.utils.data import DataLoader

import torch.nn.functional as F
import torch.optim as optim  # (可有可无)



batch_size = 64
transform = transforms.Compose([
    transforms.ToTensor(),
    transforms.Normalize((0.1307,), (0.3081,))
])

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

test_dataset = datasets.MNIST(root='../dataset/mnist/',
                              train=False,
                              download=True,
                              transform=transform)

test_loader = DataLoader(dataset=test_dataset,
                         shuffle=False,
                         batch_size=batch_size,
                         )


class Net(torch.nn.Module):
    def __init__(self):
        super().__init__() # 卷积层
        self.conv1 = torch.nn.Conv2d(in_channels=1, out_channels=10, kernel_size=5)   # 卷积层1
        self.conv2 = torch.nn.Conv2d(in_channels=10, out_channels=20, kernel_size=5)  # 卷积层2
        self.pooling = torch.nn.MaxPool2d(2)    # 池化层，没有涉及到权重，实例化一个就可以
        self.fc = torch.nn.Linear(320, 10)

    def forward(self, x):
        batch_size = x.size(0)  # 用于求维度
        x = F.relu(self.pooling(self.conv1(x)))
        x = F.relu(self.pooling(self.conv2(x)))
        x = x.view(batch_size, -1)     # 用于变成全连接
        x = self.fc(x)
        return x


model = Net()
# 选择是GPU CPU
#device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
# 表示把整个模型涉及到的权重迁移到GPU
#model.to(device)

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


def train(epoch):
    running_loss = 0.0
    for batch_index, (inputs, labels) in enumerate(train_loader, 0):
        # 迁移至GPU(模型数据要在同一块显卡上)
        # inputs, labels = inputs.to(device), labels.to(device)
        y_hat = model(inputs)
        loss = criterion(y_hat, labels)

        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        running_loss += loss.item()
        if batch_size % 10 == 9:
            print('[%d, %5d] loss: %.3f' % (epoch + 1, batch_index + 1, running_loss / 300))


def test():
    correct = 0
    total = 0
    with torch.no_grad():
        for (images, labels) in test_loader:
            # images, labels = images.to(device), labels.to(device)
            outputs = model(images)
            _, pred = torch.max(outputs.data, dim=1)
            total += labels.size(0)
            correct += (pred == labels).sum().item()
    print('accuracy on test set: %d %%' % (100 * correct / total))
    return correct / total


if __name__ == '__main__':
    epoch_list = []
    acc_list = []

    for epoch in range(10):
        train(epoch)
        acc = test()
        epoch_list.append(epoch)
        acc_list.append(acc)

    plt.plot(epoch_list, acc_list)
    plt.xlabel('epoch')
    plt.ylabel('accuracy')
    plt.show()