文章目录
前些天发现了一个巨牛的人工智能学习网站,通俗易懂,风趣幽默,忍不住分享一下给大家。点击跳转到网站。
简介
LeNet模型是在1998年提出的一种图像分类模型,应用于支票或邮件编码上的手写数字的识别,也被认为是最早的卷积神经网络(CNN),为后续CNN的发展奠定了基础,作者LeCun Y也被誉为卷积神经网络之父。LeNet之后一直直到2012年的AlexNet模型在ImageNet比赛上表现优秀,使得沉寂了14年的卷积神经网络再次成为研究热点。
LeCun Y, Bottou L, Bengio Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.
LeNet模型结构如下:
- INPUT(输入层)
输入图像的尺寸为32X32,是单通道的灰色图像。 - C1(卷积层)
使用了6个大小为5×5的卷积核,步长为1,卷积后得到6张28×28的特征图。 - S2(池化层)
使用了6个2×2 的平均池化,步长为2,池化后得到6张14×14的特征图。 - C3(卷积层)
使用了16个大小为5×5的卷积核,步长为1,得到 16 张10×10的特征图。
由于是多个卷积核对应多个输入,论文中采用了如下组合方式:
- S4(池化层)
使用16个2×2的平均池化,步长为2,池化后得到16张5×5 的特征图。 - C5(卷积层)
使用120个大小为5×5的卷积核,步长为1,卷积后得到120张1×1的特征图。 - F6(全连接层)
输入维度120,输出维度是84(对应7×12 的比特图)。 - OUTPUT(输出层)
使用高斯核函数,输入维度84,输出维度是10(对应数字 0 到 9)。
; 数据集
使用torchversion内置的 MNIST
数据集,训练集大小60000,测试集大小10000,图像大小是1×28×28,包括数字0~9共10个类。官网:http://yann.lecun.com/exdb/mnist/
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
from torchvision import transforms
import torchvision
mnist_train = torchvision.datasets.MNIST(root='./datasets/',
train=True, download=True, transform=transforms.ToTensor())
mnist_test = torchvision.datasets.MNIST(root='./datasets/',
train=False, download=True, transform=transforms.ToTensor())
print(len(mnist_train), len(mnist_test))
feature, label = mnist_train[0]
print(feature.shape, label)
dataloader = DataLoader(mnist_test, batch_size=64, num_workers=0)
step = 0
writer = SummaryWriter(log_dir='runs/mnist')
for data in dataloader:
features, labels = data
writer.add_images(tag='train', img_tensor=features, global_step=step)
step += 1
writer.close()
可视化部分可参考我这篇博客:深度学习-Tensorboard可视化面板
此外,还可以使用torchversion内置的 FashionMNIST
数据集,包括衣服、包等10类图像,也是1×28×28,各60000、10000张。
模型搭建
使用Pytoch进行搭建和测试。
在第一个卷积层C1设置padding为2,因为数据集是28×28大小,原模型是32×32大小。
import torch
from torch import nn, optim
class LeNet(nn.Module):
def __init__(self) -> None:
super().__init__()
self.model = nn.Sequential(
nn.Conv2d(in_channels=1, out_channels=6, kernel_size=5, padding=2),
nn.Sigmoid(),
nn.AvgPool2d(kernel_size=2, stride=2),
nn.Conv2d(in_channels=6, out_channels=16, kernel_size=5),
nn.Sigmoid(),
nn.AvgPool2d(kernel_size=2, stride=2),
nn.Flatten(),
nn.Linear(in_features=16 * 5 * 5, out_features=120),
nn.Sigmoid(),
nn.Linear(120, 84),
nn.Sigmoid(),
nn.Linear(in_features=84, out_features=10)
)
def forward(self, x):
return self.model(x)
leNet = LeNet()
print(leNet)
模型训练
import torch
import torchvision
from torch.utils.data import DataLoader
from torchvision import transforms
from torch import nn
from torch.utils.tensorboard import SummaryWriter
class LeNet(nn.Module):
def __init__(self) -> None:
super().__init__()
self.model = nn.Sequential(
nn.Conv2d(in_channels=1, out_channels=6, kernel_size=5, padding=2),
nn.Sigmoid(),
nn.AvgPool2d(kernel_size=2, stride=2),
nn.Conv2d(in_channels=6, out_channels=16, kernel_size=5),
nn.Sigmoid(),
nn.AvgPool2d(kernel_size=2, stride=2),
nn.Flatten(),
nn.Linear(in_features=16 * 5 * 5, out_features=120),
nn.Sigmoid(),
nn.Linear(120, 84),
nn.Sigmoid(),
nn.Linear(in_features=84, out_features=10)
)
def forward(self, x):
return self.model(x)
leNet = LeNet()
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
leNet = leNet.to(device)
loss_fn = nn.CrossEntropyLoss()
loss_fn = loss_fn.to(device)
learning_rate = 1e-2
optimizer = torch.optim.Adam(leNet.parameters(), lr=learning_rate)
total_train_step = 0
epoch = 10
writer = SummaryWriter(log_dir='./runs/LeNet/')
mnist_train = torchvision.datasets.MNIST(root='./datasets/',
train=True, download=True, transform=transforms.ToTensor())
mnist_test = torchvision.datasets.MNIST(root='./datasets/',
train=False, download=True, transform=transforms.ToTensor())
dataloader_train = DataLoader(mnist_train, batch_size=64, num_workers=0)
dataloader_test = DataLoader(mnist_test, batch_size=64, num_workers=0)
for i in range(epoch):
print("-----第{}轮训练开始-----".format(i + 1))
leNet.train()
train_loss = 0
for data in dataloader_train:
imgs, labels = data
imgs = imgs.to(device)
labels = labels.to(device)
outputs = leNet(imgs)
loss = loss_fn(outputs, labels)
optimizer.zero_grad()
loss.backward()
optimizer.step()
total_train_step += 1
train_loss += loss.item()
writer.add_scalar("train_loss_detail", loss.item(), total_train_step)
writer.add_scalar("train_loss_total", train_loss, i + 1)
torch.save(leNet, "./models/LeNet.pkl")
writer.close()
(插播反爬信息 )博主CSDN地址:https://wzlodq.blog.csdn.net/
由于打印了每轮各个批次64张图的损失,不同批次损失不同,所以上下震荡大,但总体仍是减少收敛的。
; 模型测试
leNet.eval()
total_test_loss = 0
total_accuracy = 0
with torch.no_grad():
for data in dataloader_test:
imgs, targets = data
imgs = imgs.to(device)
targets = targets.to(device)
outputs = leNet(imgs)
loss = loss_fn(outputs, targets)
total_test_loss = total_test_loss + loss.item()
accuracy = (outputs.argmax(1) == targets).sum()
total_accuracy = total_accuracy + accuracy
writer.add_scalar("test_loss", total_test_loss, i+1)
writer.add_scalar("test_accuracy", total_accuracy/len(mnist_test), i+1)
随着训练轮数增加,对应模型测试的损失减少并收敛。
精确率在几轮后就趋于98%以上,就是说感受到了来自98年的科技~
最后,附完整代码:
import torch
import torchvision
from torch.utils.data import DataLoader
from torchvision import transforms
from torch import nn
from torch.utils.tensorboard import SummaryWriter
class LeNet(nn.Module):
def __init__(self) -> None:
super().__init__()
self.model = nn.Sequential(
nn.Conv2d(in_channels=1, out_channels=6, kernel_size=5, padding=2),
nn.Sigmoid(),
nn.AvgPool2d(kernel_size=2, stride=2),
nn.Conv2d(in_channels=6, out_channels=16, kernel_size=5),
nn.Sigmoid(),
nn.AvgPool2d(kernel_size=2, stride=2),
nn.Flatten(),
nn.Linear(in_features=16 * 5 * 5, out_features=120),
nn.Sigmoid(),
nn.Linear(120, 84),
nn.Sigmoid(),
nn.Linear(in_features=84, out_features=10)
)
def forward(self, x):
return self.model(x)
leNet = LeNet()
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
leNet = leNet.to(device)
loss_fn = nn.CrossEntropyLoss()
loss_fn = loss_fn.to(device)
learning_rate = 1e-2
optimizer = torch.optim.Adam(leNet.parameters(), lr=learning_rate)
total_train_step = 0
epoch = 10
writer = SummaryWriter(log_dir='./runs/LeNet/')
mnist_train = torchvision.datasets.MNIST(root='./datasets/',
train=True, download=True, transform=transforms.ToTensor())
mnist_test = torchvision.datasets.MNIST(root='./datasets/',
train=False, download=True, transform=transforms.ToTensor())
dataloader_train = DataLoader(mnist_train, batch_size=64, num_workers=0)
dataloader_test = DataLoader(mnist_test, batch_size=64, num_workers=0)
for i in range(epoch):
print("-----第{}轮训练开始-----".format(i + 1))
leNet.train()
train_loss = 0
for data in dataloader_train:
imgs, labels = data
imgs = imgs.to(device)
labels = labels.to(device)
outputs = leNet(imgs)
loss = loss_fn(outputs, labels)
optimizer.zero_grad()
loss.backward()
optimizer.step()
total_train_step += 1
train_loss += loss.item()
writer.add_scalar("train_loss", loss.item(), total_train_step)
writer.add_scalar("train_loss", train_loss, i + 1)
leNet.eval()
total_test_loss = 0
total_accuracy = 0
with torch.no_grad():
for data in dataloader_test:
imgs, targets = data
imgs = imgs.to(device)
targets = targets.to(device)
outputs = leNet(imgs)
loss = loss_fn(outputs, targets)
total_test_loss = total_test_loss + loss.item()
accuracy = (outputs.argmax(1) == targets).sum()
total_accuracy = total_accuracy + accuracy
writer.add_scalar("test_loss", total_test_loss, i+1)
writer.add_scalar("test_accuracy", total_accuracy/len(mnist_test), i+1)
torch.save(leNet, "./models/LeNet.pkl")
writer.close()
原创不易,请勿转载(本不富裕的访问量雪上加霜 )
博主首页:https://wzlodq.blog.csdn.net/
来都来了,不评论两句吗👀
如果文章对你有帮助,记得一键三连❤Original: https://blog.csdn.net/qq_45034708/article/details/128319241
Author: 吾仄lo咚锵
Title: 深度学习-LeNet(第一个卷积神经网络)
原创文章受到原创版权保护。转载请注明出处:https://www.johngo689.com/777068/
转载文章受原作者版权保护。转载请注明原作者出处!