P5运动鞋识别学习笔记

最新推荐文章于 2024-09-16 20:12:46 发布

BOOKLOVERc

最新推荐文章于 2024-09-16 20:12:46 发布

阅读量379

点赞数 4

文章标签：学习笔记

本文链接：https://blog.csdn.net/BOOKLOVERc/article/details/136357359

版权

🍨 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
🚀 文章来源：K同学的学习圈子

目标：完成拔高训练，之后能够通过调整学习率进行调参，提高识别与预测的准确率

一、前期准备

设置GPU

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

导入数据

import os,PIL,random,pathlib

data_dir = './data5/'
data_dir = pathlib.Path(data_dir)

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

再次强化：

第一步：使用pathlib.Path()函数将字符串类型的文件夹路径转换为pathlib.Path对象。
第二步：使用glob()方法获取data_dir路径下的所有文件路径，并以列表形式存储在data_paths中。
第三步：通过split()函数对data_paths中的每个文件路径执行分割操作，获得各个文件所属的类别名称，并存储在classeNames中
第四步：打印classeNames列表，显示每个文件所属的类别名

数据集图片的处理

# 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。
])

test_transform = 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=[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。
])

train_dataset = datasets.ImageFolder("./5-data/train/",transform=train_transforms)
test_dataset  = datasets.ImageFolder("./5-data/test/",transform=train_transforms)

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)

二、构建简单的CNN网络

import torch.nn.functional as F

class Model(nn.Module):
    def __init__(self):
        super(Model, self).__init__()
        self.conv1=nn.Sequential(
            nn.Conv2d(3, 12, kernel_size=5, padding=0), # 12*220*220
            nn.BatchNorm2d(12),
            nn.ReLU())
        
        self.conv2=nn.Sequential(
            nn.Conv2d(12, 12, kernel_size=5, padding=0), # 12*216*216
            nn.BatchNorm2d(12),
            nn.ReLU())
        
        self.pool3=nn.Sequential(
            nn.MaxPool2d(2))                              # 12*108*108
        
        self.conv4=nn.Sequential(
            nn.Conv2d(12, 24, kernel_size=5, padding=0), # 24*104*104
            nn.BatchNorm2d(24),
            nn.ReLU())
        
        self.conv5=nn.Sequential(
            nn.Conv2d(24, 24, kernel_size=5, padding=0), # 24*100*100
            nn.BatchNorm2d(24),
            nn.ReLU())
        
        self.pool6=nn.Sequential(
            nn.MaxPool2d(2))                              # 24*50*50

        self.dropout = nn.Sequential(
            nn.Dropout(0.2))
        
        self.fc=nn.Sequential(
            nn.Linear(24*50*50, len(classeNames)))
        
    def forward(self, x):
        
        batch_size = x.size(0)
        x = self.conv1(x)  # 卷积-BN-激活
        x = self.conv2(x)  # 卷积-BN-激活
        x = self.pool3(x)  # 池化
        x = self.conv4(x)  # 卷积-BN-激活
        x = self.conv5(x)  # 卷积-BN-激活
        x = self.pool6(x)  # 池化
        x = self.dropout(x)
        x = x.view(batch_size, -1)  # flatten 变成全连接网络需要的输入 (batch, 24*50*50) ==> (batch, -1), -1 此处自动算出的是24*50*50
        x = self.fc(x)
       
        return x

device = "cuda" if torch.cuda.is_available() else "cpu"
print("Using {} device".format(device))

model = Model().to(device)
model

三、训练模型

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

3. 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

Epoch: 1, Train_acc:51.2%, Train_loss:0.769, Test_acc:56.6%, Test_loss:0.692, Lr:1.00E-04
Epoch: 2, Train_acc:61.8%, Train_loss:0.679, Test_acc:55.3%, Test_loss:0.703, Lr:1.00E-04
Epoch: 3, Train_acc:64.7%, Train_loss:0.635, Test_acc:69.7%, Test_loss:0.603, Lr:9.20E-05
Epoch: 4, Train_acc:67.7%, Train_loss:0.589, Test_acc:71.1%, Test_loss:0.602, Lr:9.20E-05
Epoch: 5, Train_acc:73.5%, Train_loss:0.556, Test_acc:72.4%, Test_loss:0.589, Lr:8.46E-05
Epoch: 6, Train_acc:75.3%, Train_loss:0.517, Test_acc:71.1%, Test_loss:0.526, Lr:8.46E-05
Epoch: 7, Train_acc:75.9%, Train_loss:0.521, Test_acc:71.1%, Test_loss:0.531, Lr:7.79E-05
Epoch: 8, Train_acc:79.1%, Train_loss:0.479, Test_acc:69.7%, Test_loss:0.522, Lr:7.79E-05
Epoch: 9, Train_acc:80.5%, Train_loss:0.464, Test_acc:73.7%, Test_loss:0.528, Lr:7.16E-05
Epoch:10, Train_acc:83.3%, Train_loss:0.441, Test_acc:76.3%, Test_loss:0.545, Lr:7.16E-05
Epoch:11, Train_acc:82.9%, Train_loss:0.432, Test_acc:75.0%, Test_loss:0.502, Lr:6.59E-05
Epoch:12, Train_acc:84.5%, Train_loss:0.427, Test_acc:75.0%, Test_loss:0.539, Lr:6.59E-05
Epoch:13, Train_acc:83.7%, Train_loss:0.414, Test_acc:77.6%, Test_loss:0.535, Lr:6.06E-05
Epoch:14, Train_acc:85.1%, Train_loss:0.411, Test_acc:78.9%, Test_loss:0.500, Lr:6.06E-05
Epoch:15, Train_acc:86.9%, Train_loss:0.387, Test_acc:82.9%, Test_loss:0.491, Lr:5.58E-05
Epoch:16, Train_acc:86.3%, Train_loss:0.393, Test_acc:80.3%, Test_loss:0.479, Lr:5.58E-05
Epoch:17, Train_acc:88.0%, Train_loss:0.375, Test_acc:82.9%, Test_loss:0.529, Lr:5.13E-05
Epoch:18, Train_acc:89.0%, Train_loss:0.359, Test_acc:81.6%, Test_loss:0.487, Lr:5.13E-05
Epoch:19, Train_acc:88.6%, Train_loss:0.357, Test_acc:81.6%, Test_loss:0.485, Lr:4.72E-05
Epoch:20, Train_acc:89.8%, Train_loss:0.362, Test_acc:81.6%, Test_loss:0.481, Lr:4.72E-05
Epoch:21, Train_acc:90.0%, Train_loss:0.349, Test_acc:81.6%, Test_loss:0.477, Lr:4.34E-05
Epoch:22, Train_acc:88.6%, Train_loss:0.355, Test_acc:82.9%, Test_loss:0.464, Lr:4.34E-05
Epoch:23, Train_acc:90.4%, Train_loss:0.341, Test_acc:81.6%, Test_loss:0.458, Lr:4.00E-05
Epoch:24, Train_acc:90.4%, Train_loss:0.339, Test_acc:81.6%, Test_loss:0.472, Lr:4.00E-05
Epoch:25, Train_acc:90.4%, Train_loss:0.336, Test_acc:81.6%, Test_loss:0.477, Lr:3.68E-05
Epoch:26, Train_acc:91.4%, Train_loss:0.330, Test_acc:80.3%, Test_loss:0.437, Lr:3.68E-05
Epoch:27, Train_acc:91.0%, Train_loss:0.331, Test_acc:81.6%, Test_loss:0.452, Lr:3.38E-05
Epoch:28, Train_acc:93.6%, Train_loss:0.308, Test_acc:81.6%, Test_loss:0.438, Lr:3.38E-05
Epoch:29, Train_acc:90.8%, Train_loss:0.310, Test_acc:82.9%, Test_loss:0.427, Lr:3.11E-05
Epoch:30, Train_acc:91.0%, Train_loss:0.318, Test_acc:81.6%, Test_loss:0.494, Lr:3.11E-05
Epoch:31, Train_acc:92.6%, Train_loss:0.306, Test_acc:81.6%, Test_loss:0.461, Lr:2.86E-05
Epoch:32, Train_acc:92.8%, Train_loss:0.302, Test_acc:81.6%, Test_loss:0.456, Lr:2.86E-05
Epoch:33, Train_acc:93.2%, Train_loss:0.301, Test_acc:81.6%, Test_loss:0.436, Lr:2.63E-05
Epoch:34, Train_acc:92.2%, Train_loss:0.304, Test_acc:81.6%, Test_loss:0.449, Lr:2.63E-05
Epoch:35, Train_acc:94.2%, Train_loss:0.294, Test_acc:81.6%, Test_loss:0.448, Lr:2.42E-05
Epoch:36, Train_acc:94.4%, Train_loss:0.296, Test_acc:81.6%, Test_loss:0.447, Lr:2.42E-05
Epoch:37, Train_acc:92.2%, Train_loss:0.300, Test_acc:81.6%, Test_loss:0.440, Lr:2.23E-05
Epoch:38, Train_acc:92.8%, Train_loss:0.298, Test_acc:81.6%, Test_loss:0.457, Lr:2.23E-05
Epoch:39, Train_acc:92.8%, Train_loss:0.293, Test_acc:81.6%, Test_loss:0.464, Lr:2.05E-05
Epoch:40, Train_acc:93.8%, Train_loss:0.284, Test_acc:81.6%, Test_loss:0.460, Lr:2.05E-05
Done

四、结果可视化

五.指定图片进行预测

from PIL import Image 

classes = list(train_dataset.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='./data5/test/adidas/31.jpg', 
                  model=model, 
                  transform=train_transforms, 
                  classes=classes)

预测结果是：adidas

保存并加载模型

# 模型保存
PATH = './model5.pth'  # 保存的参数文件名
torch.save(model.state_dict(), PATH)

# 将参数加载到model当中
model.load_state_dict(torch.load(PATH, map_location = device))