在目標(biāo)檢測(cè)網(wǎng)絡(luò)里加注意力機(jī)制已經(jīng)是很常見(jiàn)的了，顧名思義，注意力機(jī)制是指在全局圖像中獲得重點(diǎn)關(guān)注的目標(biāo)，常用的注意力機(jī)制有SE、CA、ECA、CBAM、GAM、NAM等。

1、SE模塊

論文：https://arxiv.org/pdf/1709.01507.pdf

參考：CV領(lǐng)域常用的注意力機(jī)制模塊（SE、CBAM）_學(xué)學(xué)沒(méi)完的博客-CSDN博客_se注意力機(jī)制

?SE模塊主要包括Squeeze和Excitation兩個(gè)部分

Squeeze是Global pooling，對(duì)特征進(jìn)行壓縮；

Excitation是通過(guò)兩層全連接結(jié)構(gòu)得到feature map中每個(gè)通道的權(quán)值，并將加權(quán)后的feature map作為下一層網(wǎng)絡(luò)的輸入。

在ECA的論文中表示，SE結(jié)構(gòu)的降維操作對(duì)通道注意有副作用。

class SELayer(nn.Module):
    def __init__(self, channel, reduction=16):
        super(SELayer, self).__init__()
        self.avg_pool = nn.AdaptiveAvgPool2d(1)
        self.fc = nn.Sequential(
            nn.Linear(channel, channel // reduction, bias=False),
            nn.ReLU(inplace=True),
            nn.Linear(channel // reduction, channel, bias=False),
            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.expand_as(x)

2、CA模塊（Coordinate attention）

論文：https://arxiv.org/abs/2103.02907

參考：CA(Coordinate attention) 注意力機(jī)制 - 知乎 (zhihu.com)

CVPR 2021 | 即插即用！ CA：新注意力機(jī)制，助力分類(lèi)/檢測(cè)/分割漲點(diǎn)！_Amusi（CVer）的博客-CSDN博客

CVPR 2021 | 即插即用！ CA：新注意力機(jī)制，助力分類(lèi)/檢測(cè)/分割漲點(diǎn)！_Amusi（CVer）的博客-CSDN博客CA對(duì)寬度和高度兩個(gè)方向分別全局平均池化，分別獲得在寬度和高度兩個(gè)方向的特征圖，然后將兩個(gè)方向的特征圖concat，然后送入共享卷積將維度降為C/r，再通過(guò)批量歸一化處理和激活函數(shù)后得到特征圖。

import torch
from torch import nn
 
 
class CA_Block(nn.Module):
    def __init__(self, channel, h, w, reduction=16):
        super(CA_Block, self).__init__()
 
        self.h = h
        self.w = w
 
        self.avg_pool_x = nn.AdaptiveAvgPool2d((h, 1))
        self.avg_pool_y = nn.AdaptiveAvgPool2d((1, w))
 
        self.conv_1x1 = nn.Conv2d(in_channels=channel, out_channels=channel//reduction, kernel_size=1, stride=1, bias=False)
 
        self.relu = nn.ReLU()
        self.bn = nn.BatchNorm2d(channel//reduction)
 
        self.F_h = nn.Conv2d(in_channels=channel//reduction, out_channels=channel, kernel_size=1, stride=1, bias=False)
        self.F_w = nn.Conv2d(in_channels=channel//reduction, out_channels=channel, kernel_size=1, stride=1, bias=False)
 
        self.sigmoid_h = nn.Sigmoid()
        self.sigmoid_w = nn.Sigmoid()
 
    def forward(self, x):
 
        x_h = self.avg_pool_x(x).permute(0, 1, 3, 2)
        x_w = self.avg_pool_y(x)
 
        x_cat_conv_relu = self.relu(self.conv_1x1(torch.cat((x_h, x_w), 3)))
 
        x_cat_conv_split_h, x_cat_conv_split_w = x_cat_conv_relu.split([self.h, self.w], 3)
 
        s_h = self.sigmoid_h(self.F_h(x_cat_conv_split_h.permute(0, 1, 3, 2)))
        s_w = self.sigmoid_w(self.F_w(x_cat_conv_split_w))
 
        out = x * s_h.expand_as(x) * s_w.expand_as(x)
 
        return out
 
 
if __name__ == '__main__':
    x = torch.randn(1, 16, 128, 64)    # b, c, h, w
    ca_model = CA_Block(channel=16, h=128, w=64)
    y = ca_model(x)
    print(y.shape)

?3、ECA模塊

論文：(PDF) ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks (researchgate.net)

參考：注意力機(jī)制（SE、Coordinate Attention、CBAM、ECA，SimAM）、即插即用的模塊整理_吳大炮的博客-CSDN博客_se注意力機(jī)制

ECA首先通過(guò)全局平均池化，然后利用卷積進(jìn)行特征提取，實(shí)現(xiàn)跨通道的交互。

4、CBAM模塊

論文：?[1807.06521] CBAM：卷積塊注意模塊 (arxiv.org)

參考：注意力機(jī)制之《CBAM: Convolutional Block Attention Module》論文閱讀_落櫻彌城的博客-CSDN博客

?

CBAM模塊分為channel-wise attention和spatial attention，通道注意力和SE結(jié)構(gòu)相同，只是加了一個(gè)maxpooling，中間共享一個(gè)MLP，最后將兩部分的輸出相加經(jīng)過(guò)sigmoid。

空間注意力使用平均池化和最大池化對(duì)輸入特征層進(jìn)行通道壓縮，在使用卷積操作。

class ChannelAttention(nn.Module):
    def __init__(self, in_planes, ratio=16):
        super(ChannelAttention, self).__init__()
        self.avg_pool = nn.AdaptiveAvgPool2d(1)
        self.max_pool = nn.AdaptiveMaxPool2d(1)

        self.fc1   = nn.Conv2d(in_planes, in_planes // ratio, 1, bias=False)
        self.relu1 = nn.ReLU()
        self.fc2   = nn.Conv2d(in_planes // ratio, in_planes, 1, bias=False)

        self.sigmoid = nn.Sigmoid()

    def forward(self, x):
        avg_out = self.fc2(self.relu1(self.fc1(self.avg_pool(x))))
        max_out = self.fc2(self.relu1(self.fc1(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__()

        assert kernel_size in (3, 7), 'kernel size must be 3 or 7'
        padding = 3 if kernel_size == 7 else 1

        self.conv1 = nn.Conv2d(2, 1, kernel_size, padding=padding, 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.conv1(x)
        return self.sigmoid(x)

5、GAM模塊

論文：https://paperswithcode.com/paper/global-attention-mechanism-retain-information

GAM注意力機(jī)制分為兩個(gè)模塊：CAM和SAM，通道注意是學(xué)習(xí)不同通道的權(quán)值，并用權(quán)值對(duì)不同通道進(jìn)行多重劃分，空間注意關(guān)注目標(biāo)在圖像上的位置信息，并通過(guò)空間特征的加權(quán)選擇性的聚焦每個(gè)空間的特征。

?通道注意力模塊首先重新排列圖像三維信息，然后通過(guò)MLP來(lái)放大跨維通道空間，如圖6所示。在空間注意子模塊中，使用兩個(gè)卷積層進(jìn)行空間信息融合，如圖7所示，這樣使通道更能關(guān)注空間信息。

import torch.nn as nn
import torch
 
 
class GAM_Attention(nn.Module):
    def __init__(self, in_channels, out_channels, rate=4):
        super(GAM_Attention, self).__init__()
 
        self.channel_attention = nn.Sequential(
            nn.Linear(in_channels, int(in_channels / rate)),
            nn.ReLU(inplace=True),
            nn.Linear(int(in_channels / rate), in_channels)
        )
 
        self.spatial_attention = nn.Sequential(
            nn.Conv2d(in_channels, int(in_channels / rate), kernel_size=7, padding=3),
            nn.BatchNorm2d(int(in_channels / rate)),
            nn.ReLU(inplace=True),
            nn.Conv2d(int(in_channels / rate), out_channels, kernel_size=7, padding=3),
            nn.BatchNorm2d(out_channels)
        )
 
    def forward(self, x):
        b, c, h, w = x.shape
        x_permute = x.permute(0, 2, 3, 1).view(b, -1, c)
        x_att_permute = self.channel_attention(x_permute).view(b, h, w, c)
        x_channel_att = x_att_permute.permute(0, 3, 1, 2)
 
        x = x * x_channel_att
 
        x_spatial_att = self.spatial_attention(x).sigmoid()
        out = x * x_spatial_att
 
        return out
 
 
if __name__ == '__main__':
    x = torch.randn(1, 64, 32, 48)
    b, c, h, w = x.shape
    net = GAM_Attention(in_channels=c, out_channels=c)
    y = net(x)
 

?6、NAM模塊

論文：https://arxiv.org/abs/2111.12419

參考：https://cloud.tencent.com/developer/article/1909196

NAM采用CBAM的模塊整合，重新設(shè)計(jì)了通道和空間注意子模塊。在通道注意模塊中使用批歸一化中的比例因子。并且將其也運(yùn)用到空間維度，來(lái)衡量像素的重要性。

文章來(lái)源地址http://www.zghlxwxcb.cn/news/detail-456725.html

import torch.nn as nn
import torch
from torch.nn import functional as F


# 具體流程可以參考圖1，通道注意力機(jī)制
class Channel_Att(nn.Module):
    def __init__(self, channels, t=16):
        super(Channel_Att, self).__init__()
        self.channels = channels
        self.bn2 = nn.BatchNorm2d(self.channels, affine=True)


    def forward(self, x):
        residual = x
        x = self.bn2(x)
        # 式2的計(jì)算，即Mc的計(jì)算
        weight_bn = self.bn2.weight.data.abs() / torch.sum(self.bn2.weight.data.abs())
        x = x.permute(0, 2, 3, 1).contiguous()
        x = torch.mul(weight_bn, x)
        x = x.permute(0, 3, 1, 2).contiguous()
        x = torch.sigmoid(x) * residual #
        
        return x


class Att(nn.Module):Yichao Liu, 2 months ago: ? Add files via upload
    def __init__(self, channels,shape, out_channels=None, no_spatial=True):
        super(Att, self).__init__()
        self.Channel_Att = Channel_Att(channels)
  
    def forward(self, x):
        x_out1=self.Channel_Att(x)
 
        return x_out1  

到了這里，關(guān)于學(xué)習(xí)筆記1——常用的注意力機(jī)制（即插即用）的文章就介紹完了。如果您還想了解更多內(nèi)容，請(qǐng)?jiān)谟疑辖撬阉鱐OY模板網(wǎng)以前的文章或繼續(xù)瀏覽下面的相關(guān)文章，希望大家以后多多支持TOY模板網(wǎng)！