GoogLeNet框架

import torch  
import torch.nn as nn  
import torch.nn.functional as F  
from torch.nn.modules.module import T  
from torch.utils.hooks import RemovableHandle  
  
  
  
class InceptionModule(nn.Module):  
 def __init__(self, in_channels, ch1x1, ch3x3red, ch3x3, ch5x5red, ch5x5, pool_proj): super().__init__() # 分支1：1x1卷积  
 self.branch1 = nn.Conv2d(in_channels, ch1x1, kernel_size=1) # 分支2：1x1 -> 3x3卷积  
 self.branch2 = nn.Sequential( nn.Conv2d(in_channels, ch3x3red, kernel_size=1), nn.Conv2d(ch3x3red, ch3x3, kernel_size=3, padding=1) ) # 分支3：1x1 -> 5x5 卷积  
 self.branch3 = nn.Sequential( nn.Conv2d(in_channels, ch5x5red, kernel_size=1), nn.Conv2d(ch5x5red, ch5x5, kernel_size=5, padding=2) ) # 分支4：3x3 池化 -> 1x1 卷积  
 self.branch4 = nn.Sequential( nn.MaxPool2d(kernel_size=3, stride=1, padding=1), nn.Conv2d(in_channels, pool_proj, kernel_size=1) ) def forward(self, x): return torch.cat([self.branch1(x), self.branch2(x), self.branch3(x), self.branch4(x)], 1)  
class GoogLeNet(nn.Module):  
 def __init__(self, num_classes=1000): super().__init__() self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3) self.maxpool1 = nn.MaxPool2d(3, stride=2, padding=1) self.conv2 = nn.Conv2d(64, 64, kernel_size=1) self.conv3 = nn.Conv2d(64, 192, kernel_size=3, padding=1) self.maxpool2 = nn.MaxPool2d(3, stride=2, padding=1) # Inception 模块堆叠  
 self.inception3a = InceptionModule(192, 64, 96, 128, 16, 32, 32) self.inception3b = InceptionModule(256, 128, 128, 192, 32, 96, 64) self.maxpool3 = nn.MaxPool2d(3, stride=2, padding=1) # 更多Inception模块  
 self.avgpool = nn.AdaptiveAvgPool2d((1, 1)) self.dropout = nn.Dropout(0.4) self.fc = nn.Linear(1024, num_classes) # 假设最终特征维度为1024  
  
 def forward(self, x): x = F.relu(self.conv1(x)) x = self.maxpool1(x) x = F.relu(self.conv2(x)) x = F.relu(self.conv3(x)) x = self.maxpool2(x) x = self.inception3a(x) x = self.inception3b(x) x = self.maxpool3(x) # 继续传递更多的模块  
 x = self.avgpool(x) x = torch.flatten(x, 1) x = self.dropout(x) x = self.fc(x) return x  

典型示例：Oxford-IIIT Pet 细粒度分类

• 目标：对 37 种宠物（猫、狗的不同品种）进行分类.
• 数据集：Oxford-IIIT Pet 数据集（包含 7,349 张图像）。

import torchvision  
import torch.optim as optim  
from torchvision import transforms  
  
# 数据预处理  
transform = transforms.Compose([  
 transforms.Resize(256), transforms.CenterCrop(224), transforms.ToTensor(), transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])])  
  
# 加载数据集  
train_set = torchvision.datasets.OxfordIIITPet(root='./data', split='trainval', download=True, transform=transform)  
train_loader = torch.utils.data.DataLoader(train_set, batch_size=32, shuffle=True)  
test_set = torchvision.datasets.OxfordIIITPet(root='./data', split='test', download=True, transform=transform)  
test_loader = torch.utils.data.DataLoader(test_set, batch_size=32, shuffle=False)  
  
# 初始化模型（调整为 37 类输出）  
model = GoogLeNet(num_classes=37)  
criterion = nn.CrossEntropyLoss()  
optimizer = optim.Adam(model.parameters(), lr=0.001)  
  
# 训练模型（使用GPU）  
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")  
model.to(device)  
  
for epoch in range(25):  
 model.train() running_loss = 0.0 for images, labels in train_loader: images, labels = images.to(device), labels.to(device) optimizer.zero_grad() outputs = model(images) loss = criterion(outputs, labels) loss.backward() optimizer.step() running_loss += loss.item() print(f'Epoch {epoch+1}, Loss: {running_loss / len(train_loader):.4f}')  
# 测试准确率  
model.eval()  
correct = 0  
total = 0  
with torch.no_grad():  
 for images, labels in test_loader: images, labels = images.to(device), labels.to(device) outputs = model(images) _, predicted = torch.max(outputs.data, 1) total += labels.size(0) correct += (predicted == labels).sum().item()print(f'Test Accuracy: {100 * correct / total:.2f}%')  

特征可视化：使用 PCA 降维展示高层特征分布：

from sklearn.decomposition import PCA  
import matplotlib.pyplot as plt  
  
features = []  
model.eval()  
with torch.no_grad():  
 for images, _ in test_loader: images = images.to(device) feat = model.avgpool(model.inception5b(images))  # 提取某层特征  
 features.append(feat.cpu().view(images.size(0), -1))features = torch.cat(features, dim=0).numpy()  
pca = PCA(n_components=2)  
reduced = pca.fit_transform(features)  
plt.scatter(reduced[:, 0], reduced[:, 1], c=test_set.targets)  
plt.show()