三分钟学会使用系列（YOLOv5）|SA注意力机制，涨点神器！

今天炼丹了吗

已于 2024-01-30 21:21:07 修改

阅读量1.1k

点赞数 18

分类专栏：深度学习文章标签： YOLO 深度学习人工智能

于 2024-01-30 21:20:56 首次发布

本文链接：https://blog.csdn.net/StopAndGoyyy/article/details/135941708

版权

深度学习专栏收录该内容

44 篇文章 17 订阅

订阅专栏

三分钟学会使用系列（YOLOv5）|SA注意力机制，涨点神器！

本文旨在让读者快速了解相关技术并运用,欢迎关注。

原文地址：《SA-NET: Shuffle attention for deep convolutional neural networks》
代码地址：《**SA-Net/models/sa_resnet.py at main · wofmanaf/SA-Net **》

1. SA注意力导读

SA-Net发表于ICASSP 2021，是由南京大学杨育彬等人提出了一种新的注意力机制：置换注意力机制（shuffle attention）。它是一种引入了特征分组与通道置换的超轻量型空间-通道注意力机制。在ImageNet与MS-COCO数据集上取得了优于SE、SGE等性能，同时具有更低的计算复杂度和参数量。
在这里插入图片描述

2. 代码

SA代码。

class sa_layer(nn.Module):
    """Constructs a Channel Spatial Group module.

    Args:
        k_size: Adaptive selection of kernel size
    """

    def __init__(self, channel, groups=64):
        super(sa_layer, self).__init__()
        self.groups = groups
        self.avg_pool = nn.AdaptiveAvgPool2d(1)
        self.cweight = Parameter(torch.zeros(1, channel // (2 * groups), 1, 1))
        self.cbias = Parameter(torch.ones(1, channel // (2 * groups), 1, 1))
        self.sweight = Parameter(torch.zeros(1, channel // (2 * groups), 1, 1))
        self.sbias = Parameter(torch.ones(1, channel // (2 * groups), 1, 1))

        self.sigmoid = nn.Sigmoid()
        self.gn = nn.GroupNorm(channel // (2 * groups), channel // (2 * groups))

    @staticmethod
    def channel_shuffle(x, groups):
        b, c, h, w = x.shape

        x = x.reshape(b, groups, -1, h, w)
        x = x.permute(0, 2, 1, 3, 4)

        # flatten
        x = x.reshape(b, -1, h, w)

        return x

    def forward(self, x):
        b, c, h, w = x.shape

        x = x.reshape(b * self.groups, -1, h, w)
        x_0, x_1 = x.chunk(2, dim=1)

        # channel attention
        xn = self.avg_pool(x_0)
        xn = self.cweight * xn + self.cbias
        xn = x_0 * self.sigmoid(xn)

        # spatial attention
        xs = self.gn(x_1)
        xs = self.sweight * xs + self.sbias
        xs = x_1 * self.sigmoid(xs)

        # concatenate along channel axis
        out = torch.cat([xn, xs], dim=1)
        out = out.reshape(b, -1, h, w)

        out = self.channel_shuffle(out, 2)
        return out

3. 使用教程

以YOLOv5为例，加入SA注意力机制。

# YOLOv5 🚀 by Ultralytics, AGPL-3.0 license

# Parameters
nc: 80  # number of classes
depth_multiple: 0.33  # model depth multiple
width_multiple: 0.25  # layer channel multiple
anchors:
  - [10,13, 16,30, 33,23]  # P3/8
  - [30,61, 62,45, 59,119]  # P4/16
  - [116,90, 156,198, 373,326]  # P5/32

# YOLOv5 v6.0 backbone
backbone:
  # [from, number, module, args]
  [[-1, 1, Conv, [64, 6, 2, 2]],  # 0-P1/2
   [-1, 1, Conv, [128, 3, 2]],  # 1-P2/4
   [-1, 3, C3, [128]],
   [-1, 1, Conv, [256, 3, 2]],  # 3-P3/8
   [-1, 6, C3, [256]],
   [-1, 1, Conv, [512, 3, 2]],  # 5-P4/16
   [-1, 9, C3, [512]],
   [-1, 1, Conv, [1024, 3, 2]],  # 7-P5/32
   [-1, 3, C3, [1024]],
   [ -1, 1, sa_layer, []],
   [-1, 1, SPPF, [1024, 5]],  # 10
  ]

# YOLOv5 v6.0 head
head:
  [[-1, 1, Conv, [512, 1, 1]],
   [-1, 1, nn.Upsample, [None, 2, 'nearest']],
   [[-1, 6], 1, Concat, [1]],  # cat backbone P4
   [-1, 3, C3, [512, False]],  # 14

   [-1, 1, Conv, [256, 1, 1]],
   [-1, 1, nn.Upsample, [None, 2, 'nearest']],
   [[-1, 4], 1, Concat, [1]],  # cat backbone P3
   [-1, 3, C3, [256, False]],  # 18 (P3/8-small)

   [-1, 1, Conv, [256, 3, 2]],
   [[-1, 14], 1, Concat, [1]],  # cat head P4
   [-1, 3, C3, [512, False]],  # 21 (P4/16-medium)

   [-1, 1, Conv, [512, 3, 2]],
   [[-1, 11], 1, Concat, [1]],  # cat head P5
   [-1, 3, C3, [1024, False]],  # 24 (P5/32-large)


   [[18, 21, 24], 1, Detect, [nc, anchors]],  # Detect(P3, P4, P5)
  ]