CBAM: Convolutional Block Attention Module

刺骨的言语ヽ痛彻心扉 2022-09-16 01:25 74阅读 0赞

### CBAM: Convolutional Block Attention Module ###

### 文章目录 ###

*   *   *  CBAM: Convolutional Block Attention Module
         *   *  参考
             *  个人理解
             *  Channel Attention
             *  Spatial Attention
             *  如何融入ResBlock中？
             *  效果如何？
             *  实现

#### 参考 ####

*  [【注意力机制】CBAM详解][CBAM]
 *  [CBAM–卷积层中的注意力模块][CBAM 1]
 *  [attention-module][]

#### 个人理解 ####

*  由于懒得系统介绍所以就长话短说，个人理解CBAM就是给Feature Map在不同维度加权重(channel和HxW)，通过改变不同channel和区域的值，从而使得网络更加关注某些区域，从而实现性能上的提升，文中用的Res50为例，且经过大佬们的验证，确实有效。
 *  大概的样子就长这样：
    
     *  ![image-20211014114409009.png][]

#### Channel Attention ####

*  它长这样
    
     *  ![image-20211014114703154][]
     *  别的都好理解，就是把Feature Map分两路走，分别作MaxPool和AvgPool（注意是全局的MaxPool和AvgPool），然后进入MLP（这里下面会说明），之后相加再激活就成了
     *  然后一开始我没太理解MLP，参考了这篇[【注意力机制】CBAM详解][CBAM]，现在我是这么理解的它的结构的，MaxPool/AvgPool的输入进入MLP之后，首先做一次1x1xC的全连接，输出C，也就是第一个长条进过一层FC，出来就是中间的较短的条，接下来就是的第二层全连接（也就是第二根长条），和第一层基本一样
        
         *  对于这个结构的作用我的理解是，给Pooling后的结果一个非线性变换，使它们在channel纬度上的特征被可以被提取出来，从而实现表示channel重要程度的功能
 *  公式的话就是下面这个，就直接截过来了，应该和图片比较对于，需要提的是$\\delta $表示的是sigmoid函数
    
     *  ![image-20211014115737643][]

#### Spatial Attention ####

*  它长这样
    
     *  ![image-20211014115949479.png][]
     *  这个就跟公式一起介绍好了，属于是比较直白，看下面的公式，我们就channel维度对Feature Map作全局Pooling后得到HxWx2的输出，此时再用7x7的卷积撸一遍，得到的Feature Map做一次sigmoid就是输出的结果了，这里需要注意的是只有一个7x7卷积核，因为HxW方向上的注意力只需要一层就够了
     *  ![image-20211014120038183][]

#### 如何融入ResBlock中？ ####

*  ResBlock的内部直接嵌入就行了，其实我的猜想是既然是个通用结构，理论上放在任何位置都OK，即你放俩ResBlock之间也不会有啥问题。
    
     *  ![image-20211014120510253][]

#### 效果如何？ ####

*  ![image-20211014120814500][]
 *  我的评价是少量参数增加+卓越的性能提升，且个人猜测参数量的增加应该主要是在channel 那里的FC处

#### 实现 ####

*  这个是抄这篇的实现：[CBAM–卷积层中的注意力模块][CBAM 1]
 *  官方代码是这份是BAM+CBAM都有：[attention-module][]

import torch
    import math
    import torch.nn as nn
    import torch.nn.functional as F
    
    class BasicConv(nn.Module):
        def __init__(self, in_planes, out_planes, kernel_size, stride=1, padding=0, dilation=1, groups=1, relu=True, bn=True, bias=False):
            super(BasicConv, self).__init__()
            self.out_channels = out_planes
            self.conv = nn.Conv2d(in_planes, out_planes, kernel_size=kernel_size, stride=stride, padding=padding, dilation=dilation, groups=groups, bias=bias)
            self.bn = nn.BatchNorm2d(out_planes,eps=1e-5, momentum=0.01, affine=True) if bn else None
            self.relu = nn.ReLU() if relu else None
    
        def forward(self, x):
            x = self.conv(x)
            if self.bn is not None:
                x = self.bn(x)
            if self.relu is not None:
                x = self.relu(x)
            return x
    
    class Flatten(nn.Module):
        def forward(self, x):
            return x.view(x.size(0), -1)
    
    class ChannelGate(nn.Module):
        def __init__(self, gate_channels, reduction_ratio=16, pool_types=['avg', 'max']):
            super(ChannelGate, self).__init__()
            self.gate_channels = gate_channels
            self.mlp = nn.Sequential(
                Flatten(),
                nn.Linear(gate_channels, gate_channels // reduction_ratio),
                nn.ReLU(),
                nn.Linear(gate_channels // reduction_ratio, gate_channels)
                )
            self.pool_types = pool_types
        def forward(self, x):
            channel_att_sum = None
            for pool_type in self.pool_types:
                if pool_type=='avg':
                    avg_pool = F.avg_pool2d( x, (x.size(2), x.size(3)), stride=(x.size(2), x.size(3)))
                    channel_att_raw = self.mlp( avg_pool )
                elif pool_type=='max':
                    max_pool = F.max_pool2d( x, (x.size(2), x.size(3)), stride=(x.size(2), x.size(3)))
                    channel_att_raw = self.mlp( max_pool )
                elif pool_type=='lp':
                    lp_pool = F.lp_pool2d( x, 2, (x.size(2), x.size(3)), stride=(x.size(2), x.size(3)))
                    channel_att_raw = self.mlp( lp_pool )
                elif pool_type=='lse':
                    # LSE pool only
                    lse_pool = logsumexp_2d(x)
                    channel_att_raw = self.mlp( lse_pool )
    
                if channel_att_sum is None:
                    channel_att_sum = channel_att_raw
                else:
                    channel_att_sum = channel_att_sum + channel_att_raw
    
            scale = F.sigmoid( channel_att_sum ).unsqueeze(2).unsqueeze(3).expand_as(x)
            return x * scale
    
    def logsumexp_2d(tensor):
        tensor_flatten = tensor.view(tensor.size(0), tensor.size(1), -1)
        s, _ = torch.max(tensor_flatten, dim=2, keepdim=True)
        outputs = s + (tensor_flatten - s).exp().sum(dim=2, keepdim=True).log()
        return outputs
    
    class ChannelPool(nn.Module):
        def forward(self, x):
            return torch.cat( (torch.max(x,1)[0].unsqueeze(1), torch.mean(x,1).unsqueeze(1)), dim=1 )
    
    class SpatialGate(nn.Module):
        def __init__(self):
            super(SpatialGate, self).__init__()
            kernel_size = 7
            self.compress = ChannelPool()
            self.spatial = BasicConv(2, 1, kernel_size, stride=1, padding=(kernel_size-1) // 2, relu=False)
        def forward(self, x):
            x_compress = self.compress(x)
            x_out = self.spatial(x_compress)
            scale = F.sigmoid(x_out) # broadcasting
            return x * scale
    
    class CBAM(nn.Module):
        def __init__(self, gate_channels, reduction_ratio=16, pool_types=['avg', 'max'], no_spatial=False):
            super(CBAM, self).__init__()
            self.ChannelGate = ChannelGate(gate_channels, reduction_ratio, pool_types)
            self.no_spatial=no_spatial
            if not no_spatial:
                self.SpatialGate = SpatialGate()
        def forward(self, x):
            x_out = self.ChannelGate(x)
            if not self.no_spatial:
                x_out = self.SpatialGate(x_out)
            return x_out

[CBAM]: https://blog.csdn.net/Roaddd/article/details/114646354
[CBAM 1]: https://zhuanlan.zhihu.com/p/96966170
[attention-module]: https://github.com/Jongchan/attention-module
[image-20211014114409009.png]: https://gitee.com/coconut_milk_candy/image-res/raw/master/blog/image-20211014114409009.png
[image-20211014114703154]: https://gitee.com/coconut_milk_candy/image-res/raw/master/blog/image-20211014114703154.png
[image-20211014115737643]: https://gitee.com/coconut_milk_candy/image-res/raw/master/blog/image-20211014115737643.png
[image-20211014115949479.png]: https://gitee.com/coconut_milk_candy/image-res/raw/master/blog/image-20211014115949479.png
[image-20211014120038183]: /images/20220828/339d11c9778e489591fad9af0a3a59ac.png
[image-20211014120510253]: /images/20220828/2e0a0b2e5b2a41dd8a7ceccb472fd4ee.png
[image-20211014120814500]: https://gitee.com/coconut_milk_candy/image-res/raw/master/blog/image-20211014120814500.png