一、完整代码
import torch
import torch.nn as nn
import torchvision.transforms as transforms
import torchvision
from torchvision import transforms, datasets
import os,PIL,pathlib,warnings
import matplotlib.pyplot as plt
# 隐藏警告
import warnings
warnings.filterwarnings("ignore") #忽略警告信息
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
device
#导入数据
import os,PIL,random,pathlib
data_dir = r'D:\编程文件\数据库\p6VGG人脸识别'
data_dir = pathlib.Path(data_dir)
data_paths = list(data_dir.glob('*'))
classeNames = [str(path).split("\\")[4] for path in data_paths]
print(classeNames)
# 关于transforms.Compose的更多介绍可以参考:https://blog.csdn.net/qq_38251616/article/details/124878863
train_transforms = transforms.Compose([
transforms.Resize([224, 224]), # 将输入图片resize成统一尺寸
# transforms.RandomHorizontalFlip(), # 随机水平翻转
transforms.ToTensor(), # 将PIL Image或numpy.ndarray转换为tensor,并归一化到[0,1]之间
transforms.Normalize( # 标准化处理-->转换为标准正太分布(高斯分布),使模型更容易收敛
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] 从数据集中随机抽样计算得到的。
])
total_data = datasets.ImageFolder(r"D:\编程文件\数据库\p6VGG人脸识别",transform=train_transforms)
print(total_data)
print(total_data.class_to_idx)
#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
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)
if __name__ == '__main__':
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
#VGG16模型
from torchvision.models import vgg16
device = "cuda" if torch.cuda.is_available() else "cpu"
print("Using {} device".format(device))
# 加载预训练模型,并且对模型进行微调
model = vgg16(pretrained=True).to(device) # 加载预训练的vgg16模型
for param in model.parameters():
param.requires_grad = False # 冻结模型的参数,这样子在训练的时候只训练最后一层的参数
# 修改classifier模块的第6层(即:(6): Linear(in_features=4096, out_features=2, bias=True))
# 注意查看我们下方打印出来的模型
model.classifier._modules['6'] = nn.Linear(4096, len(classeNames)) # 修改vgg16模型中最后一层全连接层,输出目标类别个数
model.to(device)
print(model)
# 训练循环
def train(dataloader, model, loss_fn, optimizer):
size = len(dataloader.dataset) # 训练集的大小
num_batches = len(dataloader) # 批次数目, (size/batch_size,向上取整)
train_loss, train_acc = 0, 0 # 初始化训练损失和正确率
for X, y in dataloader: # 获取图片及其标签
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,计算二者差值即为损失
# 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
#测试函数
def test(dataloader, model, loss_fn):
size = len(dataloader.dataset) # 测试集的大小
num_batches = len(dataloader) # 批次数目, (size/batch_size,向上取整)
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
#设置动态学习率
# def adjust_learning_rate(optimizer, epoch, start_lr):
# # 每 2 个epoch衰减到原来的 0.98
# lr = start_lr * (0.92 ** (epoch // 2))
# for param_group in optimizer.param_groups:
# param_group['lr'] = lr
learn_rate = 1e-4 # 初始学习率
# optimizer = torch.optim.SGD(model.parameters(), lr=learn_rate)
# 调用官方动态学习率接口时使用
lambda1 = lambda epoch: 0.92 ** (epoch // 4)
optimizer = torch.optim.SGD(model.parameters(), lr=learn_rate)
scheduler = torch.optim.lr_scheduler.LambdaLR(optimizer, lr_lambda=lambda1) #选定调整方法
#正式训练
import copy
loss_fn = nn.CrossEntropyLoss() # 创建损失函数
epochs = 40
train_loss = []
train_acc = []
test_loss = []
test_acc = []
best_acc = 0 # 设置一个最佳准确率,作为最佳模型的判别指标
if __name__ == '__main__':
for epoch in range(epochs):
# 更新学习率(使用自定义学习率时使用)
# adjust_learning_rate(optimizer, epoch, learn_rate)
model.train()
epoch_train_acc, epoch_train_loss = train(train_dl, model, loss_fn, optimizer)
scheduler.step() # 更新学习率(调用官方动态学习率接口时使用)
model.eval()
epoch_test_acc, epoch_test_loss = test(test_dl, model, loss_fn)
# 保存最佳模型到 best_model
if epoch_test_acc > best_acc:
best_acc = epoch_test_acc
best_model = copy.deepcopy(model)
train_acc.append(epoch_train_acc)
train_loss.append(epoch_train_loss)
test_acc.append(epoch_test_acc)
test_loss.append(epoch_test_loss)
# 获取当前的学习率
lr = optimizer.state_dict()['param_groups'][0]['lr']
template = ('Epoch:{:2d}, Train_acc:{:.1f}%, Train_loss:{:.3f}, Test_acc:{:.1f}%, Test_loss:{:.3f}, Lr:{:.2E}')
print(template.format(epoch + 1, epoch_train_acc * 100, epoch_train_loss,
epoch_test_acc * 100, epoch_test_loss, lr))
# 保存最佳模型到文件中
PATH = r'D:\编程文件\数据库\p6_best_model.pth' # 保存的参数文件名
torch.save(model.state_dict(), PATH)
print('Done')
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()
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) # 展示预测的图片
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照片
predict_one_image(image_path=r'D:/编程文件/数据库/p6VGG人脸识别/Angelina Jolie/001_fe3347c0.jpg',model=model,transform=train_transforms,classes=classes)
#模型评估
best_model.eval()
epoch_test_acc, epoch_test_loss = test(test_dl, best_model, loss_fn)
print(epoch_test_acc, epoch_test_loss)
# 查看是否与我们记录的最高准确率一致
print(epoch_test_acc)
二、调参过程
learn_rate = 1e-4 动态学习率(4、92);batch_size = 32;epochs = 40;SGD
(初始参数)
learn_rate = 1e-1 动态学习率(4、92);batch_size = 128;epochs = 40;SGD
learn_rate = 1e-2 动态学习率(4、92);batch_size = 100;epochs = 40;SGD
learn_rate = 5e-3 动态学习率(3、90);batch_size = 80;epochs = 40;SGD
三、结论与收获
1 . VGG-16
- VGG有三种结构,分别是VGG11、GG16和VGG19,他们并没有本质上的区别,只是网络深度不一样。
VGG11包含11个隐藏层(8个卷积层和3个全连接层)
VGG16包含了16个隐藏层(13个卷积层和3个全连接层)
VGG19包含了19个隐藏层(16个卷积层和3个全连接层) - VGG-16(Visual Geometry Group-16)是一种深度卷积神经网络(CNN)架构,使用可重复使用的卷积块来构造深度神经网络,用于图像分类和对象识别任务。
VGG-16之所以备受关注,是因为它在ImageNet图像识别竞赛中取得了很好的成绩,展示了其在大规模图像识别任务中的有效性。 - 以下是VGG-16的主要特点:
(1)深度:由13个卷积层和3个全连接层组成。
(2)卷积层的设计:3x3的卷积核(33Conv);步长为1,同时在卷积层之后都接有ReLU激活函数。
(3)池化层:2x2的最大池化层(22 MaxPool)
(4)全连接层:有3个全连接层,最后一个全连接层输出与类别数相对应的向量,用于进行分类。
2. 几种经典卷积神经网络
LeNet | AlexNet | VGG | |
---|---|---|---|
卷积层 | 3*3Conv | 更大的核窗口和步长,因为图片更大了 | 3*3Conv |
激活函数 | sigmoid | ReLu会减缓梯度消失 | ReLu |
池化层 | 2*2; 允许像素往一边移一点点 | 更大的池化窗口3*3; 往左移一下,右移一下都ok | 2*2 |
池化类型 | AvgPooling | MaxPooling | MaxPooling |
总结 | 2卷积+池化层,2全连接层 | 更大更深的LeNet; Dropout,数据增强;快 | 更大更深的AlexNet; 重复的VGG块;慢 |
LeNet : 2层卷积+3层全连接
AlexNet : 5层卷积+3层全连接
VGG11 : 8层卷积+3层全连接
VGG16 : 13层卷积+3层全连接
VGG19 : 16层卷积+3层全连接
ResNet(残差神经网络): 最大的优势就是引入了Shortcut这个支路
3. 结论
VGG使用可重复使用的卷积快来构建深度卷积神经网络
不同的卷积快个数和超参数可以得到不同复杂度的变种
训练loss一直下降,测试loss不降:过拟合啦!
高宽减半,通道数翻倍