pytorch实现猴痘病识别

•   本文为365天深度学习训练营 中的学习记录博客
• 参考文章地址： 365天深度学习训练营-第P4周：猴痘病识别
• 作者：K同学啊

一、前期准备

1.设置GPU

import torch
from torch import nn
import torchvision
from torchvision import transforms,datasets,models
import matplotlib.pyplot as plt
import os,PIL,pathlib

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

device(type='cuda')

2.导入数据

data_dir = './houdou/'
data_dir = pathlib.Path(data_dir)

data_paths = list(data_dir.glob('*'))
classNames = [str(path).split('\\')[1] for path in data_paths]
classNames

train_transforms = transforms.Compose([
    transforms.Resize([224,224]),# resize输入图片
    transforms.ToTensor(), # 将PIL Image或numpy.ndarray转换成tensor
    transforms.Normalize(
        mean = [0.485, 0.456, 0.406],
        std = [0.229,0.224,0.225]) # 从数据集中随机抽样计算得到
])

total_data = datasets.ImageFolder(data_dir,transform=train_transforms)
total_data

Dataset ImageFolder
    Number of datapoints: 2142
    Root location: houdou
    StandardTransform
Transform: Compose(
               Resize(size=[224, 224], interpolation=PIL.Image.BILINEAR)
               ToTensor()
               Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
           )

3.数据集划分

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_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)

4.数据可视化

imgs, labels = next(iter(train_dl))
imgs.shape

torch.Size([32, 3, 224, 224])

import numpy as np

 # 指定图片大小，图像大小为20宽、5高的绘图(单位为英寸inch)
plt.figure(figsize=(20, 5)) 
for i, imgs in enumerate(imgs[:20]):
    npimg = imgs.numpy().transpose((1,2,0))
    npimg = npimg * np.array((0.229, 0.224, 0.225)) + np.array((0.485, 0.456, 0.406))
    npimg = npimg.clip(0, 1)
    # 将整个figure分成2行10列，绘制第i+1个子图。
    plt.subplot(2, 10, i+1)
    plt.imshow(npimg)
    plt.axis('off')

for X,y in test_dl:
    print('Shape of X [N, C, H, W]:', X.shape)
    print('Shape of y:', y.shape)
    break

Shape of X [N, C, H, W]: torch.Size([32, 3, 224, 224])
Shape of y: torch.Size([32])

二、构建简单的CNN网络

# import torch.nn.functional as F

# num_classes = 4  # 图片的类别数

# class Network_bn(nn.Module):
#      def __init__(self):
#         super().__init__()
#          # 特征提取网络
#         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.conv3 = nn.Conv2d(in_channels=12, out_channels=24, kernel_size=5, stride=1, padding=0)
#         self.bn3 = nn.BatchNorm2d(24) 
#         self.conv4 = nn.Conv2d(in_channels=24, out_channels=24, kernel_size=5, stride=1, padding=0)
#         self.bn4 = nn.BatchNorm2d(24)  
#         # 分类网络
#         self.fc1 = nn.Linear(24*50*50,num_classes)
#      # 前向传播
#      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.bn3(self.conv3(x)))
#         x = F.relu(self.bn4(self.conv4(x)))
#         x = self.pool(x)
#         x = x.view(-1,24*50*50)
#         x = self.fc1(x)
       
#         return x
    
# model = Network_bn().to(device)
# model

2.1 迁移学习

2.1.1 调用resnet18预训练模型、冻结参数 

feature_extract = True

# 冻结参数
def set_parameter_requires_grad(model, feature_extracting):
    if feature_extracting:
        for param in model.parameters():
            param.requires_grad = False


# 修改输出层
def initialize_model(num_classes, feature_extract, use_pretrained=True):
    
    model_ft = models.resnet18(pretrained=use_pretrained)
    set_parameter_requires_grad(model_ft, feature_extract)
    
    num_ftrs = model_ft.fc.in_features
    model_ft.fc = nn.Linear(num_ftrs, num_classes)
                            
    input_size = 32

    return model_ft, input_size

model_ft, input_size = initialize_model(2, feature_extract, use_pretrained=True)
model_ft = model_ft.to(device)
model_ft

略

2.1.2 取出输出层参数 

取出输出层参数  后面用于训练更新 

# 设置训练哪些层
params_to_update = model_ft.parameters()
print("Params to learn:")
if feature_extract: # 自己只训练输出层
    params_to_update = []
    for name,param in model_ft.named_parameters():
        if param.requires_grad == True:
            params_to_update.append(param)
            print("\t",name)
else:
    for name,param in model_ft.named_parameters():
        if param.requires_grad == True:
            print("\t",name)

Params to learn:
	 fc.weight
	 fc.bias

三、训练模型

3.1 设置超参数

动态学习率 

# 优化器设置
optimizer = torch.optim.Adam(params_to_update, lr=1e-4)#要训练啥参数，你来定
scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=5, gamma=0.92)#学习率每7个epoch衰减成原来的1/10
loss_fn = nn.CrossEntropyLoss()

# 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
        
# optimizer = torch.optim.Adam(params_to_update,lr=1e-4)

3.2 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，一共900张图片
    num_batches = len(dataloader)   # 批次数目，29（900/32）

    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

3.3 编写测试函数

def test (dataloader, model, loss_fn):
    size        = len(dataloader.dataset)  # 测试集的大小，一共10000张图片
    num_batches = len(dataloader)          # 批次数目，8（255/32=8，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

3.4 正式训练

3.4.1 训练输出层

epochs     = 20
train_loss = []
train_acc  = []
test_loss  = []
test_acc   = []
best_acc = 0
filename='checkpoint.pth'


for epoch in range(epochs):
    model_ft.train()
    epoch_train_acc, epoch_train_loss = train(train_dl, model_ft, loss_fn, optimizer)
    
    scheduler.step()#学习率衰减
    
    model_ft.eval()
    epoch_test_acc, epoch_test_loss = test(test_dl, model_ft, loss_fn)
    
    # 保存最优模型
    if epoch_test_acc > best_acc:
        best_acc = epoch_test_acc
        state = {
            'state_dict': model_ft.state_dict(),#字典里key就是各层的名字，值就是训练好的权重
            'best_acc': best_acc,
            'optimizer' : optimizer.state_dict(),
        }
        torch.save(state, filename)
        
        
    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')
print('best_acc：',best_acc)

Epoch: 1, Train_acc:54.6%, Train_loss:0.707, Test_acc:54.8%，Test_loss:0.677
Epoch: 2, Train_acc:61.2%, Train_loss:0.648, Test_acc:64.8%，Test_loss:0.637
Epoch: 3, Train_acc:67.3%, Train_loss:0.615, Test_acc:67.8%，Test_loss:0.603
Epoch: 4, Train_acc:70.5%, Train_loss:0.589, Test_acc:72.5%，Test_loss:0.583
Epoch: 5, Train_acc:73.6%, Train_loss:0.567, Test_acc:75.8%，Test_loss:0.554
Epoch: 6, Train_acc:75.0%, Train_loss:0.544, Test_acc:77.2%，Test_loss:0.561
Epoch: 7, Train_acc:76.7%, Train_loss:0.530, Test_acc:80.0%，Test_loss:0.528
Epoch: 8, Train_acc:78.4%, Train_loss:0.517, Test_acc:80.4%，Test_loss:0.510
Epoch: 9, Train_acc:78.5%, Train_loss:0.503, Test_acc:80.0%，Test_loss:0.503
Epoch:10, Train_acc:78.8%, Train_loss:0.499, Test_acc:81.6%，Test_loss:0.492
Epoch:11, Train_acc:80.2%, Train_loss:0.488, Test_acc:81.8%，Test_loss:0.477
Epoch:12, Train_acc:79.7%, Train_loss:0.478, Test_acc:81.1%，Test_loss:0.477
Epoch:13, Train_acc:81.7%, Train_loss:0.471, Test_acc:82.1%，Test_loss:0.468
Epoch:14, Train_acc:81.4%, Train_loss:0.460, Test_acc:82.8%，Test_loss:0.462
Epoch:15, Train_acc:81.7%, Train_loss:0.459, Test_acc:82.8%，Test_loss:0.454
Epoch:16, Train_acc:81.3%, Train_loss:0.455, Test_acc:82.8%，Test_loss:0.440
Epoch:17, Train_acc:82.7%, Train_loss:0.447, Test_acc:83.7%，Test_loss:0.440
Epoch:18, Train_acc:82.2%, Train_loss:0.449, Test_acc:82.1%，Test_loss:0.435
Epoch:19, Train_acc:83.1%, Train_loss:0.433, Test_acc:83.2%，Test_loss:0.439
Epoch:20, Train_acc:82.9%, Train_loss:0.432, Test_acc:82.5%，Test_loss:0.430
Done

3.4.2 训练所有层

for param in model_ft.parameters():
    param.requires_grad = True

# 再继续训练所有的参数，学习率调小一点
optimizer = torch.optim.Adam(model_ft.parameters(), lr=1e-4)
scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=5, gamma=0.92)

# 损失函数
criterion = nn.CrossEntropyLoss()

# 加载之前训练好的权重参数
checkpoint = torch.load(filename)
best_acc = checkpoint['best_acc']
model_ft.load_state_dict(checkpoint['state_dict'])

epochs     = 20
train_loss = []
train_acc  = []
test_loss  = []
test_acc   = []
best_acc = 0
filename='best.pth'


for epoch in range(epochs):
    model_ft.train()
    epoch_train_acc, epoch_train_loss = train(train_dl, model_ft, loss_fn, optimizer)
    
    scheduler.step()#学习率衰减
    
    model_ft.eval()
    epoch_test_acc, epoch_test_loss = test(test_dl, model_ft, loss_fn)
    
    # 保存最优模型
    if epoch_test_acc > best_acc:
        best_acc = epoch_test_acc
        state = {
            'state_dict': model_ft.state_dict(),#字典里key就是各层的名字，值就是训练好的权重
            'best_acc': best_acc,
            'optimizer' : optimizer.state_dict(),
        }
        torch.save(state, filename)
        
        
    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')
print('best_acc：',best_acc)

Epoch: 1, Train_acc:90.7%, Train_loss:0.229, Test_acc:93.0%，Test_loss:0.171
Epoch: 2, Train_acc:97.7%, Train_loss:0.067, Test_acc:96.7%，Test_loss:0.097
Epoch: 3, Train_acc:98.1%, Train_loss:0.052, Test_acc:95.3%，Test_loss:0.142
Epoch: 4, Train_acc:99.4%, Train_loss:0.024, Test_acc:97.0%，Test_loss:0.078
Epoch: 5, Train_acc:99.2%, Train_loss:0.026, Test_acc:97.4%，Test_loss:0.077
Epoch: 6, Train_acc:99.8%, Train_loss:0.009, Test_acc:97.2%，Test_loss:0.085
Epoch: 7, Train_acc:99.7%, Train_loss:0.014, Test_acc:97.0%，Test_loss:0.114
Epoch: 8, Train_acc:99.8%, Train_loss:0.006, Test_acc:97.4%，Test_loss:0.070
Epoch: 9, Train_acc:99.6%, Train_loss:0.011, Test_acc:97.4%，Test_loss:0.080
Epoch:10, Train_acc:99.5%, Train_loss:0.014, Test_acc:95.3%，Test_loss:0.145
Epoch:11, Train_acc:99.6%, Train_loss:0.012, Test_acc:95.8%，Test_loss:0.145
Epoch:12, Train_acc:99.6%, Train_loss:0.010, Test_acc:97.2%，Test_loss:0.081
Epoch:13, Train_acc:99.8%, Train_loss:0.008, Test_acc:96.3%，Test_loss:0.131
Epoch:14, Train_acc:99.8%, Train_loss:0.005, Test_acc:96.0%，Test_loss:0.106
Epoch:15, Train_acc:100.0%, Train_loss:0.002, Test_acc:96.5%，Test_loss:0.099
Epoch:16, Train_acc:99.9%, Train_loss:0.002, Test_acc:96.3%，Test_loss:0.105
Epoch:17, Train_acc:100.0%, Train_loss:0.001, Test_acc:97.0%，Test_loss:0.085
Epoch:18, Train_acc:100.0%, Train_loss:0.001, Test_acc:96.7%，Test_loss:0.105
Epoch:19, Train_acc:100.0%, Train_loss:0.001, Test_acc:97.0%，Test_loss:0.097
Epoch:20, Train_acc:99.9%, Train_loss:0.002, Test_acc:97.4%，Test_loss:0.077
Done

四、结果可视化

加载训练好的模型

model_ft, input_size = initialize_model(2, feature_extract, use_pretrained=True)

# GPU模式
model_ft = model_ft.to(device)

# 保存文件的名字
filename='best.pth'

# 加载模型
checkpoint = torch.load(filename)
best_acc = checkpoint['best_acc']
model_ft.load_state_dict(checkpoint['state_dict'])

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()

测试模型

train_on_gpu = True

# 得到一个batch的测试数据
imgs, labels = next(iter(train_dl))

# 进行预测
model_ft.eval()

if train_on_gpu:
    output = model_ft(imgs.cuda())
else:
    output = model_ft(imgs)


# 获得预测结果(概率最大的)
_, preds_tensor = torch.max(output, 1)

preds = np.squeeze(preds_tensor.numpy()) if not train_on_gpu else np.squeeze(preds_tensor.cpu().numpy())
preds

array([1, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 0, 1, 0, 0,
       0, 1, 1, 0, 1, 0, 0, 1, 0, 0], dtype=int64)

import numpy as np

 # 指定图片大小，图像大小为20宽、5高的绘图(单位为英寸inch)
plt.figure(figsize=(20, 10)) 
for idx, imgs in enumerate(imgs[:10]):
    #ax = fig.add_subplot(rows, columns, idx+1, xticks=[], yticks=[])
    npimg = imgs.numpy().transpose((1,2,0))
    npimg = npimg * np.array((0.229, 0.224, 0.225)) + np.array((0.485, 0.456, 0.406))
    npimg = npimg.clip(0, 1)
    # 将整个figure分成2行10列，绘制第i+1个子图。
    ax = plt.subplot(2, 5, idx+1)
    ax.set_title("{} ({})".format(classNames[preds[idx]], classNames[labels[idx]]),
                 color=("green" if classNames[preds[idx]]==classNames[labels[idx]] else "red"))
    plt.imshow(npimg)
    plt.axis('off')