cbam代码

CBAM是一种用于增强卷积神经网络性能的注意力机制。

python
import torch
import torch.nn as nn
import torch.nn.functional as F

class ChannelAttention(nn.Module):
    def __init__(self, in_channels, ratio=16):
        super(ChannelAttention, self).__init__()
        self.avg_pool = nn.AdaptiveAvgPool2d(1)
        self.max_pool = nn.AdaptiveMaxPool2d(1)
        self.fc = nn.Sequential(
            nn.Conv2d(in_channels, in_channels // ratio, 1, bias=False),
            nn.ReLU(),
            nn.Conv2d(in_channels // ratio, in_channels, 1, bias=False)
        )
        self.sigmoid = nn.Sigmoid()

    def forward(self, x):
        avg_out = self.fc(self.avg_pool(x))
        max_out = self.fc(self.max_pool(x))
        out = avg_out + max_out
        return self.sigmoid(out)

class SpatialAttention(nn.Module):
    def __init__(self, kernel_size=7):
        super(SpatialAttention, self).__init__()
        self.conv = nn.Conv2d(2, 1, kernel_size, padding=(kernel_size-1)//2, bias=False)
        self.sigmoid = nn.Sigmoid()

    def forward(self, x):
        avg_out = torch.mean(x, dim=1, keepdim=True)
        max_out, _ = torch.max(x, dim=1, keepdim=True)
        x = torch.cat([avg_out, max_out], dim=1)
        x = self.conv(x)
        return self.sigmoid(x)

class CBAM(nn.Module):
    def __init__(self, in_channels):
        super(CBAM, self).__init__()
        self.channel_attention = ChannelAttention(in_channels)
        self.spatial_attention = SpatialAttention()

    def forward(self, x):
        x_c = self.channel_attention(x) * x
        x_s = self.spatial_attention(x) * x
        return x_c + x_s

# 使用示例
class YourModel(nn.Module):
    def __init__(self):
        super(YourModel, self).__init__()
        # 在你的模型中使用CBAM模块
        self.cbam = CBAM(in_channels=64)  # 例如，64是输入通道数

        # 其他模型定义...

    def forward(self, x):
        # 其他前向传播逻辑...
        x = self.cbam(x)
        return x

# 创建模型实例
model = YourModel()

上述代码是一个简化版本，实际的CBAM模块可能会根据具体的需求进行调整。此代码使用PyTorch构建CBAM模块，并在示例模型中使用了它。确保在你的项目中根据实际需要进行调整和集成。

如果你需要更详细的 CBAM 模块的实现，可以使用

python
import torch
import torch.nn as nn
import torch.nn.functional as F

class ChannelAttention(nn.Module):
    def __init__(self, in_channels, ratio=16):
        super(ChannelAttention, self).__init__()
        self.avg_pool = nn.AdaptiveAvgPool2d(1)
        self.max_pool = nn.AdaptiveMaxPool2d(1)
        
        self.fc = nn.Sequential(
            nn.Conv2d(in_channels, in_channels // ratio, 1, bias=False),
            nn.ReLU(),
            nn.Conv2d(in_channels // ratio, in_channels, 1, bias=False)
        )
        
        self.sigmoid = nn.Sigmoid()

    def forward(self, x):
        avg_out = self.fc(self.avg_pool(x))
        max_out = self.fc(self.max_pool(x))
        out = avg_out + max_out
        return self.sigmoid(out)

class SpatialAttention(nn.Module):
    def __init__(self, kernel_size=7):
        super(SpatialAttention, self).__init__()
        self.conv = nn.Conv2d(2, 1, kernel_size, padding=(kernel_size-1)//2, bias=False)
        self.sigmoid = nn.Sigmoid()

    def forward(self, x):
        avg_out = torch.mean(x, dim=1, keepdim=True)
        max_out, _ = torch.max(x, dim=1, keepdim=True)
        x = torch.cat([avg_out, max_out], dim=1)
        x = self.conv(x)
        return self.sigmoid(x)

class CBAMModule(nn.Module):
    def __init__(self, in_channels, ratio=16, kernel_size=7):
        super(CBAMModule, self).__init__()
        self.channel_attention = ChannelAttention(in_channels, ratio)
        self.spatial_attention = SpatialAttention(kernel_size)

    def forward(self, x):
        x_c = self.channel_attention(x) * x
        x_s = self.spatial_attention(x) * x
        return x_c + x_s

class SimpleCNNWithCBAM(nn.Module):
    def __init__(self, num_classes=10):
        super(SimpleCNNWithCBAM, self).__init__()
        self.conv1 = nn.Conv2d(3, 64, kernel_size=3, padding=1)
        self.cbam1 = CBAMModule(64)
        self.pool1 = nn.MaxPool2d(2, 2)
        
        self.conv2 = nn.Conv2d(64, 128, kernel_size=3, padding=1)
        self.cbam2 = CBAMModule(128)
        self.pool2 = nn.MaxPool2d(2, 2)
        
        self.fc1 = nn.Linear(128 * 8 * 8, 512)
        self.fc2 = nn.Linear(512, num_classes)

    def forward(self, x):
        x = F.relu(self.conv1(x))
        x = self.cbam1(x)
        x = self.pool1(x)
        
        x = F.relu(self.conv2(x))
        x = self.cbam2(x)
        x = self.pool2(x)
        
        x = x.view(-1, 128 * 8 * 8)
        x = F.relu(self.fc1(x))
        x = self.fc2(x)
        return x

# 创建模型实例
model = SimpleCNNWithCBAM()

这个示例中，SimpleCNNWithCBAM 是一个简单的卷积神经网络，其中使用了两个 CBAM 模块，它们分别应用于第一个和第二个卷积层。你可以根据你的需求调整模型结构和 CBAM 模块的超参数。