CBAM-通道注意力+空间注意力

class DualAttention(nn.Module):
    """双注意力模块 - 通道注意力 + 空间注意力"""
    def __init__(self, in_channels, reduction=16):
        super(DualAttention, self).__init__()
        
        # 通道注意力
        self.channel_attention = nn.Sequential(
            nn.AdaptiveAvgPool2d(1),
            nn.Conv2d(in_channels, in_channels // reduction, 1),
            nn.ReLU(inplace=True),
            nn.Conv2d(in_channels // reduction, in_channels, 1),
            nn.Sigmoid()
        )
        
        # 空间注意力
        self.spatial_attention = nn.Sequential(
            nn.Conv2d(2, 1, kernel_size=7, padding=3),
            nn.Sigmoid()
        )
    
    def forward(self, x):
        # 通道注意力
        ca = self.channel_attention(x)
        x = x * ca
        
        # 空间注意力
        avg_out = torch.mean(x, dim=1, keepdim=True)
        max_out, _ = torch.max(x, dim=1, keepdim=True)
        sa_input = torch.cat([avg_out, max_out], dim=1)
        sa = self.spatial_attention(sa_input)
        x = x * sa
        
        return x