Upload New File

人工智能系统实战第三期/实战代码/计算机视觉/CNN/network/SENet.py

+import torch
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+# 定义SE模块
+class SELayer(nn.Module):
+    def __init__(self, in_channels, reduction_ratio=16):
+        super(SELayer, self).__init__()
+        self.avg_pool = nn.AdaptiveAvgPool2d(1)
+        self.fc = nn.Sequential(
+            nn.Linear(in_channels, in_channels // reduction_ratio),
+            nn.ReLU(inplace=True),
+            nn.Linear(in_channels // reduction_ratio, in_channels),
+            nn.Sigmoid()
+        )
+
+    def forward(self, x):
+        b, c, _, _ = x.size()
+        y = self.avg_pool(x).view(b, c)
+        y = self.fc(y).view(b, c, 1, 1)
+        return x * y
+
+class SEAlexNet(nn.Module):
+    def __init__(self, num_classes=1000):
+        super(SEAlexNet, self).__init__()
+        self.features = nn.Sequential(
+            nn.Conv2d(3, 96, kernel_size=11, stride=4, padding=2),  # input[3, 224, 224]  output[96, 55, 55]
+            nn.ReLU(inplace=True),
+            nn.MaxPool2d(kernel_size=3, stride=2),  # output[96, 27, 27]
+            SELayer(96),  # 通道注意力机制
+
+            nn.Conv2d(96, 256, kernel_size=5, padding=2),  # output[256, 27, 27]
+            nn.ReLU(inplace=True),
+            nn.MaxPool2d(kernel_size=3, stride=2),  # output[256, 13, 13]
+            SELayer(256),  # 通道注意力机制
+
+            nn.Conv2d(256, 384, kernel_size=3, padding=1),  # output[384, 13, 13]
+            nn.ReLU(inplace=True),
+            SELayer(384),  # 通道注意力机制
+
+            nn.Conv2d(384, 384, kernel_size=3, padding=1),  # output[384, 13, 13]
+            nn.ReLU(inplace=True),
+            SELayer(384),  # 通道注意力机制
+
+            nn.Conv2d(384, 256, kernel_size=3, padding=1),  # output[256, 13, 13]
+            nn.ReLU(inplace=True),
+            nn.MaxPool2d(kernel_size=3, stride=2),  # output[256, 6, 6]
+        )
+        self.classifier = nn.Sequential(
+            nn.Dropout(p=0.5),
+            nn.Linear(256 * 6 * 6, 4096),
+            nn.ReLU(inplace=True),
+
+            nn.Dropout(p=0.5),
+            nn.Linear(4096, 1000),
+            nn.ReLU(inplace=True),
+
+            nn.Linear(1000, num_classes),
+        )
+
+    def forward(self, x):
+        x = self.features(x)  # 256 * [6 * 6] 6x6的图片 / 矩阵
+        x = torch.flatten(x, start_dim=1)  # 256 * 6 * 6 数列
+        x = self.classifier(x)
+        return x
+
+# 创建带有SENet模块的VGGNet实例
+# vgg_net = SEVGGNet(5)
+# # 打印网络结构
+# print(vgg_net)
\ No newline at end of file