CV中的Attention機(jī)制：簡單而有效的CBAM模塊

1. 什么是注意力機(jī)制？

• 通道注意力機(jī)制：對(duì)通道生成掩碼mask，進(jìn)行打分，代表是senet, Channel Attention Module
• 空間注意力機(jī)制：對(duì)空間進(jìn)行掩碼的生成，進(jìn)行打分，代表是Spatial Attention Module
• 混合域注意力機(jī)制：同時(shí)對(duì)通道注意力和空間注意力進(jìn)行評(píng)價(jià)打分，代表的有BAM, CBAM

2. CBAM模塊的實(shí)現(xiàn)

2.1 通道注意力機(jī)制

class ChannelAttention(nn.Module):    def __init__(self, in_planes, rotio=16):        super(ChannelAttention, self).__init__()        self.avg_pool = nn.AdaptiveAvgPool2d(1)        self.max_pool = nn.AdaptiveMaxPool2d(1)
        self.sharedMLP = nn.Sequential(            nn.Conv2d(in_planes, in_planes // ratio, 1, bias=False), nn.ReLU(),            nn.Conv2d(in_planes // rotio, in_planes, 1, bias=False))        self.sigmoid = nn.Sigmoid()
    def forward(self, x):        avgout = self.sharedMLP(self.avg_pool(x))        maxout = self.sharedMLP(self.max_pool(x))        return self.sigmoid(avgout + maxout)

2.2 空間注意力機(jī)制

class SpatialAttention(nn.Module):    def __init__(self, kernel_size=7):        super(SpatialAttention, self).__init__()        assert kernel_size in (3,7), "kernel size must be 3 or 7"        padding = 3 if kernel_size == 7 else 1
        self.conv = nn.Conv2d(2,1,kernel_size, padding=padding, bias=False)        self.sigmoid = nn.Sigmoid()
    def forward(self, x):        avgout = torch.mean(x, dim=1, keepdim=True)        maxout, _ = torch.max(x, dim=1, keepdim=True)        x = torch.cat([avgout, maxout], dim=1)        x = self.conv(x)        return self.sigmoid(x)

2.3 Convolutional bottleneck attention module

class BasicBlock(nn.Module):    expansion = 1    def __init__(self, inplanes, planes, stride=1, downsample=None):        super(BasicBlock, self).__init__()        self.conv1 = conv3x3(inplanes, planes, stride)        self.bn1 = nn.BatchNorm2d(planes)        self.relu = nn.ReLU(inplace=True)        self.conv2 = conv3x3(planes, planes)        self.bn2 = nn.BatchNorm2d(planes)        self.ca = ChannelAttention(planes)        self.sa = SpatialAttention()        self.downsample = downsample        self.stride = stride    def forward(self, x):        residual = x        out = self.conv1(x)        out = self.bn1(out)        out = self.relu(out)        out = self.conv2(out)        out = self.bn2(out)        out = self.ca(out) * out  # 廣播機(jī)制        out = self.sa(out) * out  # 廣播機(jī)制        if self.downsample is not None:            residual = self.downsample(x)        out += residual        out = self.relu(out)        return out

class cbam(nn.Module):   def __init__(self, planes)：        self.ca = ChannelAttention(planes)# planes是feature map的通道個(gè)數(shù)        self.sa = SpatialAttention()     def forward(self, x):        x = self.ca(out) * x  # 廣播機(jī)制        x = self.sa(out) * x  # 廣播機(jī)制

3. 在什么情況下可以使用？

1. 如何更有效地計(jì)算channel attention?

1. 如何更有效地計(jì)算spatial attention?

1. 如何組織這兩個(gè)部分？

4. 參考