1.设置CPU(如果有GPU就使用GPU,否则使用CPU)
>- **🍨 本文为[🔗365天深度学习训练营](https://mp.weixin.qq.com/s/Z9yL_wt7L8aPOr9Lqb1K3w) 中的学习记录博客**
>- **🍖 原作者:[K同学啊](https://mtyjkh.blog.csdn.net/)**
import torch
import torch.nn as nn
import torchvision.transforms as transforms
import torchvision
from torchvision import transforms,datasets
import os,PIL,pathlib
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
device
device(type='cpu')
2.导入数据
import os,PIL,random,pathlib
data_dir = './4-data/'
data_dir = pathlib.Path(data_dir)
data_paths = list(data_dir.glob('*'))
classeNames = [str(path).split("/")[1] for path in data_paths]
classeNames
['.DS_Store', 'Others', 'Monkeypox']
total_datadir = './4-data/'
train_transforms = transforms.Compose([
transforms.Resize([224,224]), ## 将输入图片resize成统一尺寸
transforms.ToTensor(), ## 将PIL Image或numpy.ndarray转换成tensor并归一化到[0, 1]之间
transforms.Normalize( ## 标准化处理-->转换为正态分布(高斯分布),使模型更加容易收敛
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]) ## 其中 mean=[0.485, 0.456, 0.406]与std=[0.229, 0.224, 0.225])从数据集随机抽样得到的
])
total_data = datasets.ImageFolder(total_datadir,transform=train_transforms)
total_data
Dataset ImageFolder
Number of datapoints: 2142
Root location: ./4-data/
StandardTransform
Transform: Compose(
Resize(size=[224, 224], interpolation=bilinear, max_size=None, antialias=True)
ToTensor()
Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
)
total_data.class_to_idx
{'Monkeypox': 0, 'Others': 1}
3.划分数据集
train_size = int(0.8 * len(total_data))
test_size = len(total_data) - train_size
train_dataset, test_dataset = torch.utils.data.random_split(total_data, [train_size, test_size])
train_dataset, test_dataset
(<torch.utils.data.dataset.Subset at 0x146f09090>,
<torch.utils.data.dataset.Subset at 0x146f094b0>)
train_size, test_size
(1713, 429)
batch_size = 32
train_dl = torch.utils.data.DataLoader(train_dataset,
batch_size = batch_size,
shuffle = True,
num_workers = 1)
test_dl = torch.utils.data.DataLoader(test_dataset,
batch_size = batch_size,
shuffle = True,
num_workers = 1)
for X, y in test_dl:
print("Shape of X [N, C, H, W]:", X.shape)
print("Shape of y", y.shape, y.dtype)
break
Shape of X [N, C, H, W]: torch.Size([32, 3, 224, 224])
Shape of y torch.Size([32]) torch.int64
二、构建简单的CNN网络
import torch.nn.functional as F
class Network_bn(nn.Module):
def __init__(self):
super(Network_bn, self).__init__()
self.conv1 = nn.Conv2d(in_channels = 3, out_channels = 12, kernel_size = 5, stride = 1, padding = 0)
self.bn1 = nn.BatchNorm2d(12)
self.conv2 = nn.Conv2d(in_channels = 12, out_channels = 12, kernel_size = 5, stride = 1, padding = 0)
self.bn2 = nn.BatchNorm2d(12)
self.pool = nn.MaxPool2d(2, 2)
self.conv4 = nn.Conv2d(in_channels = 12, out_channels = 24, kernel_size = 5, stride = 1, padding = 0)
self.bn4 = nn.BatchNorm2d(24)
self.conv5 = nn.Conv2d(in_channels = 24, out_channels = 24, kernel_size = 5, stride = 1, padding = 0)
self.bn5 = nn.BatchNorm2d(24)
self.fc1 = nn.Linear(24*50*50, len(classeNames))
def forward(self, x):
x = F.relu(self.bn1(self.conv1(x)))
x = F.relu(self.bn2(self.conv2(x)))
x = self.pool(x)
x = F.relu(self.bn4(self.conv4(x)))
x = F.relu(self.bn5(self.conv5(x)))
x = self.pool(x)
x = x.view(-1, 24*50*50)
x = self.fc1(x)
return x
device = "cuda" if torch.cuda.is_available() else "cpu"
print("Using {} device".format(device))
model = Network_bn().to(device)
model
Using cpu device
Network_bn(
(conv1): Conv2d(3, 12, kernel_size=(5, 5), stride=(1, 1))
(bn1): BatchNorm2d(12, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(conv2): Conv2d(12, 12, kernel_size=(5, 5), stride=(1, 1))
(bn2): BatchNorm2d(12, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(conv4): Conv2d(12, 24, kernel_size=(5, 5), stride=(1, 1))
(bn4): BatchNorm2d(24, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(conv5): Conv2d(24, 24, kernel_size=(5, 5), stride=(1, 1))
(bn5): BatchNorm2d(24, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(fc1): Linear(in_features=60000, out_features=3, bias=True)
)
三、训练模型
1.设置超参数
loss_fn = nn.CrossEntropyLoss() ## 创建损失函数
learn_rate = 1e-4 ## 学习率
opt = torch.optim.SGD(model.parameters(), lr = learn_rate)
2.编写训练函数
## 训练循环
def train(dataloader, model, loss_fn, optimizer):
size = len(dataloader.dataset)
num_batches = len(dataloader)
train_loss, train_acc = 0, 0
for X, y in dataloader:
X, y = X.to(device), y.to(device)
## 计算预测误差
pred = model(X) ## 网络输出
loss = loss_fn(pred, y) ## 计算网络输出和真实值之间的差距,targets为真实值,计算二者差即为损失
## 反向传播
optimizer.zero_grad() ## grad属性归零
loss.backward() ## 反向传播
optimizer.step() ## 每一步自动更新
## 记录acc与loss
train_acc += (pred.argmax(1) == y).type(torch.float).sum().item() ##
train_loss += loss.item()
train_acc /= size
train_loss /= num_batches
return train_acc, train_loss
3.编写测试函数
def test (dataloader, model, loss_fn):
size = len(dataloader.dataset)
num_batches = len(dataloader)
test_loss, test_acc = 0, 0
## 当不进行训练时,停止梯度更新,节省计算内存消耗
with torch.no_grad():
for imgs, target in dataloader:
imgs, target = imgs.to(device), target.to(device)
## 计算loss
target_pred = model(imgs)
loss = loss_fn(target_pred, target)
test_loss += loss.item()
test_acc += (target_pred.argmax(1) == target).type(torch.float).sum().item()
test_acc /= size
test_loss /= num_batches
return test_acc, test_loss
4.正式训练
epochs = 20
train_loss = []
train_acc = []
test_loss = []
test_acc = []
for epoch in range(epochs):
model.train()
epoch_train_acc, epoch_train_loss = train(train_dl, model, loss_fn, opt)
model.eval()
epoch_test_acc, epoch_test_loss = test(test_dl, model, loss_fn)
train_acc.append(epoch_train_acc)
train_loss.append(epoch_train_loss)
test_acc.append(epoch_test_acc)
test_loss.append(epoch_test_loss)
template = ('Epoch:{:2d}, Train_acc:{:.1f}%, Train_loss:{:.3f}, Test_acc:{:.1f}%, Test_loss:{:.3f}')
print(template.format(epoch+1, epoch_train_acc*100, epoch_train_loss, epoch_test_acc*100, epoch_test_loss))
print('Done')
Epoch: 1, Train_acc:58.9%, Train_loss:0.745, Test_acc:65.5%, Test_loss:0.637
Epoch: 2, Train_acc:67.7%, Train_loss:0.634, Test_acc:68.5%, Test_loss:0.596
Epoch: 3, Train_acc:71.8%, Train_loss:0.562, Test_acc:70.6%, Test_loss:0.571
Epoch: 4, Train_acc:74.4%, Train_loss:0.536, Test_acc:72.0%, Test_loss:0.571
Epoch: 5, Train_acc:77.6%, Train_loss:0.489, Test_acc:73.0%, Test_loss:0.553
Epoch: 6, Train_acc:80.2%, Train_loss:0.455, Test_acc:75.8%, Test_loss:0.518
Epoch: 7, Train_acc:81.7%, Train_loss:0.439, Test_acc:76.0%, Test_loss:0.500
Epoch: 8, Train_acc:81.9%, Train_loss:0.426, Test_acc:79.3%, Test_loss:0.479
Epoch: 9, Train_acc:84.6%, Train_loss:0.397, Test_acc:80.0%, Test_loss:0.460
Epoch:10, Train_acc:86.3%, Train_loss:0.371, Test_acc:80.4%, Test_loss:0.448
Epoch:11, Train_acc:87.2%, Train_loss:0.358, Test_acc:81.1%, Test_loss:0.436
Epoch:12, Train_acc:88.3%, Train_loss:0.347, Test_acc:82.5%, Test_loss:0.433
Epoch:13, Train_acc:87.9%, Train_loss:0.337, Test_acc:83.2%, Test_loss:0.413
Epoch:14, Train_acc:89.3%, Train_loss:0.323, Test_acc:83.2%, Test_loss:0.424
Epoch:15, Train_acc:89.5%, Train_loss:0.312, Test_acc:84.6%, Test_loss:0.424
Epoch:16, Train_acc:91.9%, Train_loss:0.295, Test_acc:84.1%, Test_loss:0.396
Epoch:17, Train_acc:91.2%, Train_loss:0.288, Test_acc:84.1%, Test_loss:0.387
Epoch:18, Train_acc:91.5%, Train_loss:0.279, Test_acc:83.4%, Test_loss:0.376
Epoch:19, Train_acc:92.0%, Train_loss:0.277, Test_acc:84.4%, Test_loss:0.393
Epoch:20, Train_acc:92.2%, Train_loss:0.270, Test_acc:83.7%, Test_loss:0.377
Done
四、结果可视化
import matplotlib.pyplot as plt
## 隐藏警告
import warnings
warnings.filterwarnings("ignore") ## 忽略警告信息
plt.rcParams['font.sans-serif'] = ['SimHei'] ## 用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False ## 用来正常显示负号
plt.rcParams['figure.dpi'] = 100 ## 分辨率
epochs_range = range(epochs)
plt.figure(figsize = (12, 3))
plt.subplot(1, 2, 1)
plt.plot(epochs_range, train_acc, label = 'Training Accuracy')
plt.plot(epochs_range, test_acc, label = 'Test Accuracy')
plt.legend(loc = 'lower right')
plt.title('Training and Validation Accuracy')
plt.subplot(1, 2, 2)
plt.plot(epochs_range, train_loss, label = 'Training Loss')
plt.plot(epochs_range, test_loss, label= 'Test Loss')
plt.legend(loc = 'upper right')
plt.title('Training and Validation Loss')
plt.show()
2.指定图片进行预测
from PIL import Image
classes = list(total_data.class_to_idx)
def predict_one_image(image_path, model, transform, classes):
test_img = Image.open(image_path).convert('RGB')
## plt.imshow(test_img) ## 展示预测图片
test_img = transform(test_img)
img = test_img.to(device).unsqueeze(0)
model.eval()
output = model(img)
_,pred = torch.max(output,1)
pred_class = classes[pred]
print(f'预测结果是:{pred_class}')
## 预测训练集中的某张图片
predict_one_image(image_path='./4-data/Monkeypox/M01_01_00.jpg',
model=model,
transform=train_transforms,
classes=classes)
预测结果是:Monkeypox
五、保存并加载模型
## 模型保存
PATH = './model.pth' ## 保存的参数文件名
torch.save(model.state_dict(), PATH)
## 将参数加载到model中
model.load_state_dict(torch.load(PATH, map_location=device))
<All keys matched successfully>
由于初始模型的Test_acc不是很高,所以以下将尝试一些改进:
改进1: 增加卷积核数量
import torch.nn.functional as F
class Network_bn(nn.Module):
def __init__(self):
super(Network_bn, self).__init__()
self.conv1 = nn.Conv2d(in_channels = 3, out_channels = 32, kernel_size = 5, stride = 1, padding = 0)
self.bn1 = nn.BatchNorm2d(32)
self.conv2 = nn.Conv2d(in_channels = 32, out_channels = 64, kernel_size = 5, stride = 1, padding = 0)
self.bn2 = nn.BatchNorm2d(64)
self.pool = nn.MaxPool2d(2, 2)
self.conv4 = nn.Conv2d(in_channels = 64, out_channels = 128, kernel_size = 5, stride = 1, padding = 0)
self.bn4 = nn.BatchNorm2d(128)
self.conv5 = nn.Conv2d(in_channels = 128, out_channels = 128, kernel_size = 5, stride = 1, padding = 0)
self.bn5 = nn.BatchNorm2d(128)
self.fc1 = nn.Linear(128*50*50, len(classeNames))
def forward(self, x):
x = F.relu(self.bn1(self.conv1(x)))
x = F.relu(self.bn2(self.conv2(x)))
x = self.pool(x)
x = F.relu(self.bn4(self.conv4(x)))
x = F.relu(self.bn5(self.conv5(x)))
x = self.pool(x)
x = x.view(-1, 128*50*50)
x = self.fc1(x)
return x
device = "cuda" if torch.cuda.is_available() else "cpu"
print("Using {} device".format(device))
model = Network_bn().to(device)
model
改进后:训练到epoch5发现效果很不好
原因分析:可能是增加卷积核数量后模型复杂度升高,而数据集规模不是很大,从而导致了过拟合。
改进2: 使用小卷积核(3*3)
import torch.nn.functional as F
class Network_bn(nn.Module):
def __init__(self):
super(Network_bn, self).__init__()
self.conv1 = nn.Conv2d(in_channels = 3, out_channels = 12, kernel_size = 3, stride = 1, padding = 0)
self.bn1 = nn.BatchNorm2d(12)
self.conv2 = nn.Conv2d(in_channels = 12, out_channels = 12, kernel_size = 3, stride = 1, padding = 0)
self.bn2 = nn.BatchNorm2d(12)
self.pool = nn.MaxPool2d(2, 2)
self.conv4 = nn.Conv2d(in_channels = 12, out_channels = 24, kernel_size = 3, stride = 1, padding = 0)
self.bn4 = nn.BatchNorm2d(24)
self.conv5 = nn.Conv2d(in_channels = 24, out_channels = 24, kernel_size = 3, stride = 1, padding = 0)
self.bn5 = nn.BatchNorm2d(24)
self.fc1 = nn.Linear(24*53*53, len(classeNames))
def forward(self, x):
x = F.relu(self.bn1(self.conv1(x)))
x = F.relu(self.bn2(self.conv2(x)))
x = self.pool(x)
x = F.relu(self.bn4(self.conv4(x)))
x = F.relu(self.bn5(self.conv5(x)))
x = self.pool(x)
x = x.view(-1, 24*53*53)
x = self.fc1(x)
return x
device = "cuda" if torch.cuda.is_available() else "cpu"
print("Using {} device".format(device))
model = Network_bn().to(device)
model
改进后结果:
对比之前5*5卷积核,结果没有提高。
原因分析:可能是由于模型感受野的减小,3乘3卷积核单次卷积操作中捕获的图像信息范围相较于5*5有所减小,因此模型对较大或全局特征的捕获能力不足。对于猴痘病识别这样的任务,可能某些大尺度特征对于分类比较重要,可能3乘3卷积核不足以提取这些特征。
改进3: 在原来的基础上增加dropout层:
import torch.nn.functional as F
class Network_bn(nn.Module):
def __init__(self):
super(Network_bn, self).__init__()
self.conv1 = nn.Conv2d(in_channels = 3, out_channels = 12, kernel_size = 5, stride = 1, padding = 0)
self.bn1 = nn.BatchNorm2d(12)
self.conv2 = nn.Conv2d(in_channels = 12, out_channels = 12, kernel_size = 5, stride = 1, padding = 0)
self.bn2 = nn.BatchNorm2d(12)
self.pool = nn.MaxPool2d(2, 2)
self.conv4 = nn.Conv2d(in_channels = 12, out_channels = 24, kernel_size = 5, stride = 1, padding = 0)
self.bn4 = nn.BatchNorm2d(24)
self.conv5 = nn.Conv2d(in_channels = 24, out_channels = 24, kernel_size = 5, stride = 1, padding = 0)
self.bn5 = nn.BatchNorm2d(24)
self.dropout = nn.Dropout(0.5) ## dropout层
self.fc1 = nn.Linear(24*50*50, len(classeNames))
def forward(self, x):
x = F.relu(self.bn1(self.conv1(x)))
x = F.relu(self.bn2(self.conv2(x)))
x = self.pool(x)
x = F.relu(self.bn4(self.conv4(x)))
x = F.relu(self.bn5(self.conv5(x)))
x = self.pool(x)
x = x.view(-1, 24*50*50)
## 在全连接层之前加入dropout层
x = self.dropout(x)
x = self.fc1(x)
return x
device = "cuda" if torch.cuda.is_available() else "cpu"
print("Using {} device".format(device))
model = Network_bn().to(device)
model
改进后结果:
对比最开始的模型,Test_acc并没有提高。
原因分析:有可能参数选择不好,导致过度正则化,特别是在模型规模不大时。
同时有可能模型结构比较简单,加入 Dropout 后模型的表达能力不足,因此影响了其在测试集上的表现。
改进4: 应用动态学习率
loss_fn = nn.CrossEntropyLoss() ## 创建损失函数
learn_rate = 1e-4 ## 学习率
opt = torch.optim.Adam(model.parameters(), lr = learn_rate, weight_decay=1e-4) ## 使用Adam优化器
scheduler = torch.optim.lr_scheduler.StepLR(opt, step_size=10, gamma=0.1) ## 动态学习率
改进结果:
可以发现Test_acc达到90%改进效果较为显著。
总结:
在改进模型结构时:
1.我们增加其深度即增加卷积核核个数,理论上是可以提高其测试集准确度,但是在训练集规模较小时可能会过拟合。
2.我们在尝试减小卷积核时,理论上是可以通过小的卷积核来更细致地捕捉图像的局部特征。这样有助于模型提取更多有用的细节信息,从而提高识别准确率。但是同时模型感受野会减小,从而导致模型对较大或全局特征的捕获能力不足,最终影响分类性能。
设置动态学习率,可以根据策略自动调整学习率,帮助模型更快收敛并且避免其陷入局部最优解。使用效果也是较为突出。