YOLOv8改进 - 注意力篇 - 引入EMA注意力机制

news2024/10/7 6:05:25

一、本文介绍

作为入门性篇章，这里介绍了EMA注意力在YOLOv8中的使用。包含EMA原理分析，EMA的代码、EMA的使用方法、以及添加以后的yaml文件及运行记录。

二、EMA原理分析

EMA官方论文地址：EMA文章

EMA代码：EMA代码

EMA注意力机制（高效的多尺度注意力）：通过重塑部分通道到批次维度，并将通道维度分组为多个子特征，以保留每个通道的信息并减少计算开销。EMA模块通过编码全局信息来重新校准每个并行分支中的通道权重，并通过跨维度交互来捕获像素级别的关系。

相关代码：

EMA注意力的代码，如下：

class EMA_attention(nn.Module):
    def __init__(self, channels, c2=None, factor=32):
        super(EMA_attention, self).__init__()
        self.groups = factor
        assert channels // self.groups > 0
        self.softmax = nn.Softmax(-1)
        self.agp = nn.AdaptiveAvgPool2d((1, 1))
        self.pool_h = nn.AdaptiveAvgPool2d((None, 1))
        self.pool_w = nn.AdaptiveAvgPool2d((1, None))
        self.gn = nn.GroupNorm(channels // self.groups, channels // self.groups)
        self.conv1x1 = nn.Conv2d(channels // self.groups, channels // self.groups, kernel_size=1, stride=1, padding=0)
        self.conv3x3 = nn.Conv2d(channels // self.groups, channels // self.groups, kernel_size=3, stride=1, padding=1)

    def forward(self, x):
        b, c, h, w = x.size()
        group_x = x.reshape(b * self.groups, -1, h, w)  # b*g,c//g,h,w
        x_h = self.pool_h(group_x)
        x_w = self.pool_w(group_x).permute(0, 1, 3, 2)
        hw = self.conv1x1(torch.cat([x_h, x_w], dim=2))
        x_h, x_w = torch.split(hw, [h, w], dim=2)
        x1 = self.gn(group_x * x_h.sigmoid() * x_w.permute(0, 1, 3, 2).sigmoid())
        x2 = self.conv3x3(group_x)
        x11 = self.softmax(self.agp(x1).reshape(b * self.groups, -1, 1).permute(0, 2, 1))
        x12 = x2.reshape(b * self.groups, c // self.groups, -1)  # b*g, c//g, hw
        x21 = self.softmax(self.agp(x2).reshape(b * self.groups, -1, 1).permute(0, 2, 1))
        x22 = x1.reshape(b * self.groups, c // self.groups, -1)  # b*g, c//g, hw
        weights = (torch.matmul(x11, x12) + torch.matmul(x21, x22)).reshape(b * self.groups, 1, h, w)
        return (group_x * weights.sigmoid()).reshape(b, c, h, w)

四、YOLOv8中EMA使用方法

1.YOLOv8中添加EMA模块：

首先在ultralytics/nn/modules/conv.py最后添加EMA模块的代码。

2.在conv.py的开头__all__ = 内添加EMA模块的类别名EMA_attention：

3.在同级文件夹下的__init__.py内添加EMA的相关内容：（分别是from .conv import EMA_attention ；以及在__all__内添加EMA_attention）

4.在ultralytics/nn/tasks.py进行EMA_attention注意力机制的注册，以及在YOLOv8的yaml配置文件中添加EMA_attention即可。

首先打开task.py文件，按住Ctrl+F，输入parse_model进行搜索。找到parse_model函数。在其最后一个else前面(或者直接加载含有c2f模块的注册代码内，注册方式一样)添加以下注册代码：

        if m in (EMA_attention):
            c1, c2 = ch[f], args[0]

然后，就是新建一个名为YOLOv8_EMA.yaml的配置文件：（路径:ultralytics/cfg/models/v8/YOLOv8_EMA.yaml）其中参数中nc，由自己的数据集决定。本文测试，采用的coco8数据集，有80个类别。

# Ultralytics YOLO 🚀, AGPL-3.0 license
# YOLOv8 object detection model with P3-P5 outputs. For Usage examples see https://docs.ultralytics.com/tasks/detect

# Parameters
nc: 80  # number of classes
scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call CPAM-yolov8.yaml with scale 'n'
  # [depth, width, max_channels]
  n: [0.33, 0.25, 1024]  # YOLOv8n summary: 225 layers,  3157200 parameters,  3157184 gradients,   8.9 GFLOPs
  s: [0.33, 0.50, 1024]  # YOLOv8s summary: 225 layers, 11166560 parameters, 11166544 gradients,  28.8 GFLOPs
  m: [0.67, 0.75, 768]   # YOLOv8m summary: 295 layers, 25902640 parameters, 25902624 gradients,  79.3 GFLOPs
  l: [1.00, 1.00, 512]   # YOLOv8l summary: 365 layers, 43691520 parameters, 43691504 gradients, 165.7 GFLOPs
  x: [1.00, 1.25, 512]   # YOLOv8x summary: 365 layers, 68229648 parameters, 68229632 gradients, 258.5 GFLOPs

# YOLOv8.0n backbone
backbone:
  # [from, repeats, module, args]
  - [-1, 1, Conv, [64, 3, 2]]  # 0-P1/2
  - [-1, 1, Conv, [128, 3, 2]]  # 1-P2/4
  - [-1, 3, C2f, [128, True]]
  - [-1, 1, Conv, [256, 3, 2]]  # 3-P3/8
  - [-1, 6, C2f, [256, True]]
  - [-1, 1, Conv, [512, 3, 2]]  # 5-P4/16
  - [-1, 6, C2f, [512, True]]
  - [-1, 1, Conv, [1024, 3, 2]]  # 7-P5/32
  - [-1, 3, C2f, [1024, True]]
  - [-1, 1, EMA_attention, [1024]]
  - [-1, 1, SPPF, [1024, 5]]  # 9

# YOLOv8.0n head
head:
  - [-1, 1, nn.Upsample, [None, 2, 'nearest']]
  - [[-1, 6], 1, Concat, [1]]  # cat backbone P4
  - [-1, 3, C2f, [512]]  # 12

  - [-1, 1, nn.Upsample, [None, 2, 'nearest']]
  - [[-1, 4], 1, Concat, [1]]  # cat backbone P3
  - [-1, 3, C2f, [256]]  # 15 (P3/8-small)

  - [-1, 1, Conv, [256, 3, 2]]
  - [[-1, 13], 1, Concat, [1]]  # cat head P4
  - [-1, 3, C2f, [512]]  # 18 (P4/16-medium)

  - [-1, 1, Conv, [512, 3, 2]]
  - [[-1, 10], 1, Concat, [1]]  # cat head P5
  - [-1, 3, C2f, [1024]]  # 21 (P5/32-large)

  - [[16, 19, 22], 1, Detect, [nc]]  # Detect(P3, P4, P5)

在根目录新建一个train.py文件，内容如下:

from ultralytics import YOLO
import warnings

with warnings.catch_warnings():
    warnings.simplefilter("ignore")

    model = YOLO('ultralytics/cfg/models/v8/YOLOv8_EMA.yaml')  # 从YAML建立一个新模型
    results = model.train(data='ultralytics/cfg/datasets/coco8.yaml', epochs=1,imgsz=640,optimizer="SGD")

训练输出：