3、天气识别

• 🍨 本文为🔗365天深度学习训练营 中的学习记录博客
• 🍦 参考文章：365天深度学习训练营-第P2周：彩色识别
• 🍖 原作者：K同学啊|接辅导、项目定制

 

学习要求¶

• 本地读取并加载数据。
• 测试集accuracy到达93%

学习重点¶

测试集accuracy到达95%

调用模型识别一张本地图片

 

一、前期工作准备部分¶

1、设置GPU¶

In [1]:

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

Out[1]:

device(type='cuda')

 

2、导入数据¶

本次是进行本地数据的导入，首先下载到了本地，进行云的图像读取。 导入数据主要用的是pathlib的库，这个库在python3.4之后成为标准库模块，具有跨平台、面向对象等特点，解决了传统路径与字符串不等价的问题，使得在不同操作系统之间切换简单。

🐒pathlib主要包含两个类，PurePath和Path，两个类分别有两个子类，分别用于unix系统和windows系统

下面的代码中 classname = [str(path).split("\")[3] for path in paths] 困扰我最久

• split("\")[3]指将path中的字段用\作为分隔符，[3]是选取其中第四个字段（python以0作为第一个字段）
• str(path)是将path作为字符来处理。

In [2]:

data_dir = '../data/weather_photos'
# 通过Path类创建路径对象
data_dir = pathlib.Path(data_dir)
# 获取路径下所有文件路径
paths= list(data_dir.glob('*'))
# 获取所有文件夹的名字，也就是图片类别
classname = [str(path).split("\\")[3] for path in paths]
print(classname)

# 图像transforms
# 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.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和std是从数据中随机抽样计算得到的
])
total_data = datasets.ImageFolder(data_dir, transform=train_transforms)
total_data

 

['cloudy', 'rain', 'shine', 'sunrise']

Out[2]:

Dataset ImageFolder
    Number of datapoints: 1125
    Root location: ..\data\weather_photos
    StandardTransform
Transform: Compose(
               Resize(size=[224, 224], interpolation=bilinear, max_size=None, antialias=None)
               ToTensor()
               Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
           )

 

3、划分数据集¶

In [3]:

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

Out[3]:

(<torch.utils.data.dataset.Subset at 0x231cc35aeb8>,
 <torch.utils.data.dataset.Subset at 0x231cc35ae48>)

In [4]:

train_size,test_size

Out[4]:

(900, 225)

In [5]:

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)

In [6]:

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¶

对于一般的CNN网络来说，都是由特征提取网络和分类网络构成，其中特征提取网络用于提取图片的特征，分类网络用于将图片进行分类。

In [7]:

import torch.nn.functional as F

class Network_bn(nn.Module):
    def __init__(self):
        super(Network_bn, self).__init__()
        """
        nn.Conv2d()函数：
        第一个参数（in_channels）是输入的channel数量
        第二个参数（out_channels）是输出的channel数量
        第三个参数（kernel_size）是卷积核大小
        第四个参数（stride）是步长，默认为1
        第五个参数（padding）是填充大小，默认为0
        """
        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(classname))

    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 cuda device

Out[7]:

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=4, bias=True)
)

 

三、 训练模型¶

1、设置超参数¶

In [8]:

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¶

In [9]:

# 训练循环
def train(dataloader, model, loss_fn, optimizer):
    size = len(dataloader.dataset)  # 训练集的大小，一共60000张图片
    num_batches = len(dataloader)   # 批次数目，1875（60000/32）

    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、编写测试函数¶

In [10]:

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，313（10000/32=312.5，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

 

4、正式训练¶

In [11]:

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:64.0%, Train_loss:0.949, Test_acc:62.7%，Test_loss:1.002
Epoch: 2, Train_acc:80.6%, Train_loss:0.633, Test_acc:80.9%，Test_loss:0.676
Epoch: 3, Train_acc:85.0%, Train_loss:0.551, Test_acc:74.7%，Test_loss:0.526
Epoch: 4, Train_acc:84.6%, Train_loss:0.467, Test_acc:84.9%，Test_loss:0.393
Epoch: 5, Train_acc:88.3%, Train_loss:0.420, Test_acc:87.6%，Test_loss:0.397
Epoch: 6, Train_acc:89.2%, Train_loss:0.398, Test_acc:85.8%，Test_loss:0.423
Epoch: 7, Train_acc:88.8%, Train_loss:0.378, Test_acc:88.4%，Test_loss:0.343
Epoch: 8, Train_acc:91.4%, Train_loss:0.325, Test_acc:88.9%，Test_loss:0.337
Epoch: 9, Train_acc:91.3%, Train_loss:0.305, Test_acc:87.1%，Test_loss:0.321
Epoch:10, Train_acc:91.6%, Train_loss:0.332, Test_acc:86.7%，Test_loss:0.364
Epoch:11, Train_acc:92.2%, Train_loss:0.290, Test_acc:87.6%，Test_loss:0.350
Epoch:12, Train_acc:92.1%, Train_loss:0.276, Test_acc:88.4%，Test_loss:0.354
Epoch:13, Train_acc:92.8%, Train_loss:0.257, Test_acc:86.2%，Test_loss:0.308
Epoch:14, Train_acc:92.3%, Train_loss:0.260, Test_acc:88.9%，Test_loss:0.309
Epoch:15, Train_acc:93.1%, Train_loss:0.239, Test_acc:87.1%，Test_loss:0.416
Epoch:16, Train_acc:93.8%, Train_loss:0.225, Test_acc:88.9%，Test_loss:0.330
Epoch:17, Train_acc:94.9%, Train_loss:0.217, Test_acc:88.9%，Test_loss:0.307
Epoch:18, Train_acc:94.9%, Train_loss:0.210, Test_acc:88.9%，Test_loss:0.305
Epoch:19, Train_acc:95.2%, Train_loss:0.195, Test_acc:88.9%，Test_loss:0.279
Epoch:20, Train_acc:95.0%, Train_loss:0.215, Test_acc:88.9%，Test_loss:0.365
Done

 

5、可视化¶

In [13]:

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()