【Week-P6】VGG16-好莱坞明星识别

• 本文为365天深度学习训练营 中的学习记录博客
• 原作者：K同学啊 | 接辅导、项目定制

说明：
（1）本次学习使用VGG-16模型完成，调用官方接口得到VGG-16预训练模型，然后修改classifier模块的第6层；
（2）本次学习调用官方动态学习率接口完成训练；
（3）本次学习需要调整调整参数，使test_accuracy的值达到60%或以上（当前训练40个epoch的test_accuracy是16.7%）；
（4）本次学习使用的数据集与Week T6 - 好莱坞明星识别（CNN）的数据集是一致的，共有17位好莱坞明星的脸部照片；

一、环境配置

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 sys
from datetime import datetime
print("---------------------1.配置环境------------------")
print("Start time: ", datetime.today())
print("Pytorch version: " + torch.__version__)
print("Python version: " + sys.version)

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
device

二、准备数据

2.1 打印classNames列表，显示每个文件所属的类别名称
2.2 打印归一化后的类别名称，0或1
2.3 划分数据集，划分为训练集&测试集，torch.utils.data.DataLoader()参数详解
2.4 检查数据集的shape

import os,PIL,random,pathlib
print("---------------------2.1 导入数据------------------")
data_dir = 'D:/jupyter notebook/DL-100-days/datasets/hollywood-celebraties/'
data_dir = pathlib.Path(data_dir)

data_paths  = list(data_dir.glob('*'))
classNames = [str(path).split("\\")[5] for path in data_paths]
print("classNames:", classNames)

# 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("D:/jupyter notebook/DL-100-days/datasets/hollywood-celebraties/",transform=train_transforms)
print("total_data:", total_data)
print("total_data.class_to_idx: ", total_data.class_to_idx)

print("---------------------2.2 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])
print("train_dataset: ", train_dataset)
print("test_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)
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

三、搭建网络结构

VGG-16结构说明：

● 13个卷积层（Convolutional Layer），分别用blockX_convX表示；
● 3个全连接层（Fully connected Layer），用classifier表示；
● 5个池化层（Pool layer）。

'''
调用官方VGG-16模型， 修改classifier模块的第6层
'''
print("---------------------3. 调用官方VGG-16模型，修改classifier模块的第6层------------------")
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(classNames)) # 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)  
model

四、开始训练

4.1 编写训练函数
4.2 编写测试函数
4.3 设置动态学习率（调用官方接口）
4.4 开始正式训练

print("---------------------4.1 编写训练函数------------------")
# 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

print("---------------------4.2 编写测试函数------------------")
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

print("---------------------4.3 设置动态学习率------------------")
'''
自定义动态学习率：
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)

'''
# 调用官方的动态学习率方法，下面几行与三引号注释里的代码使等价的
learn_rate = 1e-4 # 初始学习率
# 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

'''
model.train()、model.eval()训练营往期文章中有详细的介绍。

请注意观察保存最佳模型的方式，与TensorFlow2的保存方式有何异同。
'''
print("---------------------4.4 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

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 = 'D:/jupyter notebook/DL-100-days/datasets/P6_best_model.pth'  # 保存的参数文件名
torch.save(model.state_dict(), PATH)

print('Done')

Pytorch保存最佳模型的方式：

Tensorflow保存最佳模型的方式：

五、查看训练结果

5.1 绘制训练结果图
5.2 指定图片进行预测
5.3 模型评估

print("---------------------5.1 绘制训练结果图------------------")
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()

print("---------------------5.2 指定图片进行预测------------------")
from PIL import Image 

classes = list(total_data.class_to_idx)
# 预测训练集中的某张照片
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}')
    
predict_one_image(image_path='D:/jupyter notebook/DL-100-days/datasets/hollywood-celebraties/Angelina Jolie/002_8f8da10e.jpg', 
                  model=model, 
                  transform=train_transforms, 
                  classes=classes)

print("---------------------5.3 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_acc)
print("epoch_test_loss: ", epoch_test_loss)
# 查看是否与我们记录的最高准确率一致
print("epoch_test_acc: ", epoch_test_acc)

六、修改部分参数，提高test_accuracy

6.1 修改为每2个epoch学习率衰减一次，得到test_accuracy = 24.7%

lambda1 = lambda epoch: 0.92 ** (epoch // 2)

6.1 修改三个地方，得到test_accuracy = 23.9%

（1）batch_size从32修改为64；
（2）初始学习率修改为0.0002；
（3）在模型内部增加参数初始化，参考这里；

batch_size = 64
learn_rate = 0.0002 # 初始学习率
# VGG16：自己的模型
class My_VGG16(nn.Module):
    def __init__(self,num_classes=5,init_weight=True):
        super(My_VGG16, self).__init__()
        # 特征提取层
        self.features = nn.Sequential(
            nn.Conv2d(in_channels=3,out_channels=64,kernel_size=3,stride=1,padding=1),
            nn.Conv2d(in_channels=64,out_channels=64,kernel_size=3,stride=1,padding=1),
            nn.MaxPool2d(kernel_size=2,stride=2),
            nn.Conv2d(in_channels=64, out_channels=128, kernel_size=3, stride=1, padding=1),
            nn.Conv2d(in_channels=128, out_channels=128, kernel_size=3, stride=1, padding=1),
            nn.MaxPool2d(kernel_size=2, stride=2),
            nn.Conv2d(in_channels=128, out_channels=256, kernel_size=3, stride=1, padding=1),
            nn.Conv2d(in_channels=256, out_channels=256, kernel_size=3, stride=1, padding=1),
            nn.Conv2d(in_channels=256, out_channels=256, kernel_size=3, stride=1, padding=1),
            nn.MaxPool2d(kernel_size=2,stride=2),
            nn.Conv2d(in_channels=256, out_channels=512, kernel_size=3, stride=1, padding=1),
            nn.Conv2d(in_channels=512, out_channels=512, kernel_size=3, stride=1, padding=1),
            nn.Conv2d(in_channels=512, out_channels=512, kernel_size=3, stride=1, padding=1),
            nn.MaxPool2d(kernel_size=2, stride=2),
            nn.Conv2d(in_channels=512, out_channels=512, kernel_size=3, stride=1, padding=1),
            nn.Conv2d(in_channels=512, out_channels=512, kernel_size=3, stride=1, padding=1),
            nn.Conv2d(in_channels=512, out_channels=512, kernel_size=3, stride=1, padding=1),
            nn.MaxPool2d(kernel_size=2, stride=2),
        )
        # 分类层
        self.classifier = nn.Sequential(
            nn.Linear(in_features=7*7*512,out_features=4096),
            nn.ReLU(),
            nn.Dropout(0.5),
            nn.Linear(in_features=4096,out_features=4096),
            nn.ReLU(),
            nn.Dropout(0.5),
            nn.Linear(in_features=4096,out_features=len(classNames))
        )

        # 参数初始化
        if init_weight: # 如果进行参数初始化
            for m in self.modules():  # 对于模型的每一层
                if isinstance(m, nn.Conv2d): # 如果是卷积层
                    # 使用kaiming初始化
                    nn.init.kaiming_normal_(m.weight, mode="fan_out", nonlinearity="relu")
                    # 如果bias不为空，固定为0
                    if m.bias is not None:
                        nn.init.constant_(m.bias, 0)
                elif isinstance(m, nn.Linear):# 如果是线性层
                    # 正态初始化
                    nn.init.normal_(m.weight, 0, 0.01)
                    # bias则固定为0
                    nn.init.constant_(m.bias, 0)

    def forward(self,x):
        x = self.features(x)
        x = torch.flatten(x,1)
        result = self.classifier(x)
        return result
'''
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
                        
原文链接：https://blog.csdn.net/weixin_46676835/article/details/129582927
'''
model = My_VGG16()
model.to(device)  
model

训练次数不变，得到的训练结果如下：