YOLOv8优化：注意力系列篇 | ECA注意力，效果秒杀CBAM、SE-灵析社区

神机妙算

🚀🚀🚀本文改进：ECA注意力，引入到YOLOv8，多种实现方式

🚀🚀🚀ECA在不同检测领域中应用广泛

🚀🚀🚀YOLOv8改进专栏：http://t.csdnimg.cn/hGhVK

2.ECA



论文：https://arxiv.org/pdf/1910.03151.pdf

本文的贡献总结如下:

对SE模块进行了剖析，并分别证明了避免降维和适当的跨通道交互对于学习高性能和高效率的通道注意力是重要的。

在以上分析的基础上，提出了一种高效通道注意模块(ECA)，在CNN网络上提出了一种极轻量的通道注意力模块，该模块增加的模型复杂度小，提升效果显著。

在ImageNet-1K和MS COCO上的实验结果表明，本文提出的方法具有比目前最先进的CNN模型更低的模型复杂度，与此同时，本文方法却取得了非常有竞争力的结果。

       作者设计了一个高效的channel attention机制，该方法保留了原有的通道一对一权重更新，并且通过local cross-channel interaction来提升结果。此外作者还设计了一个自动调节kernel size的机制来决定交叉学习的覆盖率。通过该ECA模块，作者在几乎一样的参数上获得了分类top-1 acc 2%的提升

​


ECA（ECA-Net）是一种新的卷积神经网络结构，它的优势在于以下几点：

  1.  能够有效地捕捉长距离依赖关系。ECA-Net引入了一种新的通道注意力机制，可以对不同通道的特征进行加权，从而使网络能够更好地捕捉长距离依赖关系。
  2.  参数量小、计算效率高。ECA-Net的参数量比传统的卷积神经网络更小，同时也能够在保证准确率的前提下大大提高计算效率。
  3.  可以适应不同的输入尺寸。ECA-Net可以适应不同的输入尺寸，因此在处理不同大小的图像时具有很好的适应性。

.总的来说，ECA-Net是一种非常有前途的卷积神经网络结构，它具有良好的性能和高效的计算能力，可以在计算机视觉等领域得到广泛的应用。




​


       比较了不同的注意力方法在ImageNet数据集上的网络参数(param)，浮点运算每秒(FLOPs)，训练或推理速度(帧每秒，FPS)， Top-1/Top-5的准确性(%)。



2.ECA加入YOLOv8

2.1加入ultralytics/nn/attention/attention.py

import torch
import torch.nn as nn
from torch.nn import functional as F
 
######################  ECAAttention ####     start     ###############################
 
 
class ECAAttention(nn.Module):
    """Constructs a ECA module.
    Args:
        channel: Number of channels of the input feature map
        k_size: Adaptive selection of kernel size
    """
 
    def __init__(self, c1, k_size=3):
        super(ECAAttention, self).__init__()
        self.avg_pool = nn.AdaptiveAvgPool2d(1)
        self.conv = nn.Conv1d(1, 1, kernel_size=k_size, padding=(k_size - 1) // 2, bias=False)
        self.sigmoid = nn.Sigmoid()
 
    def forward(self, x):
        # feature descriptor on the global spatial information
        y = self.avg_pool(x)
        y = self.conv(y.squeeze(-1).transpose(-1, -2)).transpose(-1, -2).unsqueeze(-1)
        # Multi-scale information fusion
        y = self.sigmoid(y)
 
        return x * y.expand_as(x)
 
######################  ECAAttention ####     end     ###############################

2.2 修改tasks.py

首先ECAAttention进行注册


from ultralytics.nn.attention.attention import *

函数def parse_model(d, ch, verbose=True): # model_dict, input_channels(3)进行修改


        if m in (Classify, Conv, ConvTranspose, GhostConv, Bottleneck, GhostBottleneck, SPP, SPPF, DWConv, Focus,
                 BottleneckCSP, C1, C2, C2f, C3, C3TR, C3Ghost, nn.ConvTranspose2d, DWConvTranspose2d, C3x, RepC3,ECAAttention):
            c1, c2 = ch[f], args[0]

2.3 yaml实现

2.3.1 yolov8_ECAAttention.yaml

加入backbone SPPF后




# 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 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, SPPF, [1024, 5]]  # 9
  - [-1, 1, ECAAttention, [1024]]  # 10
 
# YOLOv8.0n head
head:
  - [-1, 1, nn.Upsample, [None, 2, 'nearest']]
  - [[-1, 6], 1, Concat, [1]]  # cat backbone P4
  - [-1, 3, C2f, [512]]  # 13
 
  - [-1, 1, nn.Upsample, [None, 2, 'nearest']]
  - [[-1, 4], 1, Concat, [1]]  # cat backbone P3
  - [-1, 3, C2f, [256]]  # 16 (P3/8-small)
 
  - [-1, 1, Conv, [256, 3, 2]]
  - [[-1, 13], 1, Concat, [1]]  # cat head P4
  - [-1, 3, C2f, [512]]  # 19 (P4/16-medium)
 
  - [-1, 1, Conv, [512, 3, 2]]
  - [[-1, 10], 1, Concat, [1]]  # cat head P5
  - [-1, 3, C2f, [1024]]  # 22 (P5/32-large)
 
  - [[16, 19, 22], 1, Detect, [nc]]  # Detect(P3, P4, P5)

2.3.2 yolov8_ECAAttention2.yaml

neck里的连接Detect的3个C2f结合



# 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 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, 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, ECAAttention, [256]]  # 16
 
  - [-1, 1, Conv, [256, 3, 2]]
  - [[-1, 12], 1, Concat, [1]]  # cat head P4
  - [-1, 3, C2f, [512]]  # 19 (P4/16-medium)
  - [-1, 1, ECAAttention, [512]]  # 20
 
  - [-1, 1, Conv, [512, 3, 2]]
  - [[-1, 9], 1, Concat, [1]]  # cat head P5
  - [-1, 3, C2f, [1024]]  # 23 (P5/32-large)
  - [-1, 1, ECAAttention, [1024]]  # 24
 
  - [[16, 20, 24], 1, Detect, [nc]]  # Detect(P3, P4, P5)

2.3.3 yolov8_ECAAttention3.yaml

放入neck的C2f后面



# Ultralytics YOLO 🚀, GPL-3.0 license
# YOLOv8 object detection model with P3-P5 outputs. For Usage examples see https://docs.ultralytics.com/tasks/detect
 
# Parameters
nc: 1  # number of classes
scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call 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, 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, ECAAttention, [512]]  # 13
 
  - [-1, 1, nn.Upsample, [None, 2, 'nearest']]
  - [[-1, 4], 1, Concat, [1]]  # cat backbone P3
  - [-1, 3, C2f, [256]]  # 16 (P3/8-small)
  - [-1, 1, ECAAttention, [256]]  # 17 (P5/32-large)
 
  - [-1, 1, Conv, [256, 3, 2]]
  - [[-1, 13], 1, Concat, [1]]  # cat head P4
  - [-1, 3, C2f, [512]]  # 20 (P4/16-medium)
  - [-1, 1, ECAAttention, [512]]  # 21 (P5/32-large)
 
  - [-1, 1, Conv, [512, 3, 2]]
  - [[-1, 9], 1, Concat, [1]]  # cat head P5
  - [-1, 3, C2f, [1024]]  # 24 (P5/32-large)
  - [-1, 1, ECAAttention, [1024]]  # 25 (P5/32-large)
 
  - [[17, 21, 25], 1, Detect, [nc]]  # Detect(P3, P4, P5)


阅读量：572

点赞量：0

收藏量:0