YOLO11改进-模块-引入基于部分卷积的前馈网络PCFN

2024-10-31 07:07:33

          本篇文章将介绍一个新的改进模块——PCFN（部分卷积前馈网络），并阐述如何将其应用于YOLOv11中，显著提升模型性能。首先，我们将解析PCFN的工作原理，它通过部分卷积和跨通道交互的方式来加强特征表达。随后，我们会详细说明如何将该模块与YOLOv11相结合，展示代码实现细节及其使用方法，最终展现该改进对目标检测效果的积极影响。  

1. Partial Convolution-based Feed-forward Network (PCFN)结构介绍       

        PCFN（基于部分卷积的前馈网络）： PCFN采用部分卷积来减少计算冗余，并通过跨通道交互来增强特征表示。具体地，PCFN首先通过1×1卷积进行通道交互，然后将特征分割为两部分，对其中一部分进行3×3卷积处理，最后再将两部分特征融合，生成更具代表性的输出特征。PCFN不仅能够在通道维度上实现特征融合，还能够在空间维度上对特征进行局部编码，从而进一步提高图像的特征。

2. YOLOv11与PCFN的结合

     PCFN通过跨通道和空间的交互，进一步增强网络提取的特征。PCFN的设计理念是减少计算冗余的同时，保持高效的特征融合。

        1. 本文将PCFN与C2PSA相结合，使用PCFN替换其中的FFN模块，构建C2PSA_PCFN模块。

        2. 本文将PCFN与C3K2相结合，使用PCFN替换conv模块，构建C3k2_PCFN模块。

3. Partial Convolution-based Feed-forward Network (PCFN)代码部分

import torch
import torch.nn as nn
from .block import PSABlock, C2PSA, C2f, C3, Bottleneck
from .conv import Conv


class PCFN(nn.Module):
    def __init__(self, dim, growth_rate=2.0, p_rate=0.25):
        super().__init__()
        hidden_dim = int(dim * growth_rate)
        p_dim = int(hidden_dim * p_rate)
        self.conv_0 = nn.Conv2d(dim,hidden_dim,1,1,0)
        self.conv_1 = nn.Conv2d(p_dim, p_dim ,3,1,1)

        self.act =nn.GELU()
        self.conv_2 = nn.Conv2d(hidden_dim, dim, 1, 1, 0)

        self.p_dim = p_dim
        self.hidden_dim = hidden_dim

    def forward(self, x):
        if self.training:
            x = self.act(self.conv_0(x))
            x1, x2 = torch.split(x,[self.p_dim,self.hidden_dim-self.p_dim],dim=1)
            x1 = self.act(self.conv_1(x1))
            x = self.conv_2(torch.cat([x1,x2], dim=1))
        else:
            x = self.act(self.conv_0(x))
            x[:,:self.p_dim,:,:] = self.act(self.conv_1(x[:,:self.p_dim,:,:]))
            x = self.conv_2(x)
        return x


class PSABlock_PCFN(PSABlock):

    def __init__(self, c, qk_dim =16 , pdim=32, shortcut=True) -> None:
        """Initializes the PSABlock with attention and feed-forward layers for enhanced feature extraction."""
        super().__init__(c)
        self.ffn = PCFN(c)


class C2PSA_PCFN(C2PSA):

    def __init__(self, c1, c2, n=1, e=0.5):
        """Initializes the C2PSA module with specified input/output channels, number of layers, and expansion ratio."""
        super().__init__(c1, c2)
        assert c1 == c2
        self.c = int(c1 * e)
        self.m = nn.Sequential(*(PSABlock_PCFN(self.c, qk_dim =16 , pdim=32) for _ in range(n)))


class Bottleneck_PCFN(nn.Module):
    """Standard bottleneck."""

    def __init__(self, c1, c2, shortcut=True, g=1, k=(3, 3), e=0.5):
        """Initializes a standard bottleneck module with optional shortcut connection and configurable parameters."""
        super().__init__()
        c_ = int(c2 * e)  # hidden channels
        self.cv1 = Conv(c1, c_, k[0], 1)
        self.cv2 = PCFN(c_)
        self.add = shortcut and c1 == c2

    def forward(self, x):
        """Applies the YOLO FPN to input data."""
        return x + self.cv2(self.cv1(x)) if self.add else self.cv2(self.cv1(x))

class C3k(C3):
    """C3k is a CSP bottleneck module with customizable kernel sizes for feature extraction in neural networks."""

    def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5, k=3):
        """Initializes the C3k module with specified channels, number of layers, and configurations."""
        super().__init__(c1, c2, n, shortcut, g, e)
        c_ = int(c2 * e)  # hidden channels
        # self.m = nn.Sequential(*(RepBottleneck(c_, c_, shortcut, g, k=(k, k), e=1.0) for _ in range(n)))
        self.m = nn.Sequential(*(Bottleneck_PCFN(c_, c_, shortcut, g, k=(k, k), e=1.0) for _ in range(n)))

# 在c3k=True时，使用C3k2_PCFN特征融合，为false的时候我们使用普通的Bottleneck提取特征
class C3k2_PCFN(C2f):
    """Faster Implementation of CSP Bottleneck with 2 convolutions."""

    def __init__(self, c1, c2, n=1, c3k=False, e=0.5, g=1, shortcut=True):
        """Initializes the C3k2 module, a faster CSP Bottleneck with 2 convolutions and optional C3k blocks."""
        super().__init__(c1, c2, n, shortcut, g, e)
        self.m = nn.ModuleList(
            C3k(self.c, self.c, 2, shortcut, g) if c3k else Bottleneck(self.c, self.c, shortcut, g) for _ in range(n)
        )


if __name__ == '__main__':
    TB = PCFN(256)
    #创建一个输入张量
    batch_size = 8
    input_tensor=torch.randn(batch_size, 256, 64, 64 )
    #运行模型并打印输入和输出的形状
    output_tensor =TB(input_tensor)
    print("Input shape:",input_tensor.shape)
    print("0utput shape:",output_tensor.shape)

 4. 将PCFN引入到YOLOv11中

第一: 将下面的核心代码复制到D:\bilibili\model\YOLO11\ultralytics-main\ultralytics\nn路径下，如下图所示。

第二：在task.py中导入PCFN包

第三：在task.py中的模型配置部分下面代码

        第一个改进需要修改的地方

         第二个改进需要修改的地方

第四：将模型配置文件复制到YOLOV11.YAMY文件中

        第一个改进的配置文件

# Ultralytics YOLO ????, AGPL-3.0 license
# YOLO11 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=yolo11n.yaml' will call yolo11.yaml with scale 'n'
  # [depth, width, max_channels]
  n: [0.50, 0.25, 1024] # summary: 319 layers, 2624080 parameters, 2624064 gradients, 6.6 GFLOPs
  s: [0.50, 0.50, 1024] # summary: 319 layers, 9458752 parameters, 9458736 gradients, 21.7 GFLOPs
  m: [0.50, 1.00, 512] # summary: 409 layers, 20114688 parameters, 20114672 gradients, 68.5 GFLOPs
  l: [1.00, 1.00, 512] # summary: 631 layers, 25372160 parameters, 25372144 gradients, 87.6 GFLOPs
  x: [1.00, 1.50, 512] # summary: 631 layers, 56966176 parameters, 56966160 gradients, 196.0 GFLOPs

# YOLO11n 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, 2, C3k2, [256, False, 0.25]]
  - [-1, 1, Conv, [256, 3, 2]] # 3-P3/8
  - [-1, 2, C3k2, [512, False, 0.25]]
  - [-1, 1, Conv, [512, 3, 2]] # 5-P4/16
  - [-1, 2, C3k2, [512, True]]
  - [-1, 1, Conv, [1024, 3, 2]] # 7-P5/32
  - [-1, 2, C3k2, [1024, True]]
  - [-1, 1, SPPF, [1024, 5]] # 9
  - [-1, 2, C2PSA_PCFN, [1024]] # 10

# YOLO11n head
head:
  - [-1, 1, nn.Upsample, [None, 2, "nearest"]]
  - [[-1, 6], 1, Concat, [1]] # cat backbone P4
  - [-1, 2, C3k2, [512, False]] # 13

  - [-1, 1, nn.Upsample, [None, 2, "nearest"]]
  - [[-1, 4], 1, Concat, [1]] # cat backbone P3
  - [-1, 2, C3k2, [256, False]] # 16 (P3/8-small)

  - [-1, 1, Conv, [256, 3, 2]]
  - [[-1, 13], 1, Concat, [1]] # cat head P4
  - [-1, 2, C3k2, [512, False]] # 19 (P4/16-medium)

  - [-1, 1, Conv, [512, 3, 2]]
  - [[-1, 10], 1, Concat, [1]] # cat head P5
  - [-1, 2, C3k2, [1024, True]] # 22 (P5/32-large)

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

        第二个改进的配置文件

# Ultralytics YOLO ????, AGPL-3.0 license
# YOLO11 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=yolo11n.yaml' will call yolo11.yaml with scale 'n'
  # [depth, width, max_channels]
  n: [0.50, 0.25, 1024] # summary: 319 layers, 2624080 parameters, 2624064 gradients, 6.6 GFLOPs
  s: [0.50, 0.50, 1024] # summary: 319 layers, 9458752 parameters, 9458736 gradients, 21.7 GFLOPs
  m: [0.50, 1.00, 512] # summary: 409 layers, 20114688 parameters, 20114672 gradients, 68.5 GFLOPs
  l: [1.00, 1.00, 512] # summary: 631 layers, 25372160 parameters, 25372144 gradients, 87.6 GFLOPs
  x: [1.00, 1.50, 512] # summary: 631 layers, 56966176 parameters, 56966160 gradients, 196.0 GFLOPs

# YOLO11n 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, 2, C3k2_PCFN, [256, False, 0.25]]
  - [-1, 1, Conv, [256, 3, 2]] # 3-P3/8
  - [-1, 2, C3k2_PCFN, [512, False, 0.25]]
  - [-1, 1, Conv, [512, 3, 2]] # 5-P4/16
  - [-1, 2, C3k2_PCFN, [512, True]]
  - [-1, 1, Conv, [1024, 3, 2]] # 7-P5/32
  - [-1, 2, C3k2_PCFN, [1024, True]]
  - [-1, 1, SPPF, [1024, 5]] # 9
  - [-1, 2, C2PSA, [1024]] # 10

# YOLO11n head
head:
  - [-1, 1, nn.Upsample, [None, 2, "nearest"]]
  - [[-1, 6], 1, Concat, [1]] # cat backbone P4
  - [-1, 2, C3k2_PCFN, [512, False]] # 13

  - [-1, 1, nn.Upsample, [None, 2, "nearest"]]
  - [[-1, 4], 1, Concat, [1]] # cat backbone P3
  - [-1, 2, C3k2_PCFN, [256, False]] # 16 (P3/8-small)

  - [-1, 1, Conv, [256, 3, 2]]
  - [[-1, 13], 1, Concat, [1]] # cat head P4
  - [-1, 2, C3k2_PCFN, [512, False]] # 19 (P4/16-medium)

  - [-1, 1, Conv, [512, 3, 2]]
  - [[-1, 10], 1, Concat, [1]] # cat head P5
  - [-1, 2, C3k2_PCFN, [1024, True]] # 22 (P5/32-large)

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

第五：运行成功


from ultralytics.models import NAS, RTDETR, SAM, YOLO, FastSAM, YOLOWorld

if __name__=="__main__":


    # 使用自己的YOLOv11.yamy文件搭建模型并加载预训练权重训练模型
    model = YOLO(r"D:\bilibili\model\YOLO11\ultralytics-main\ultralytics\cfg\models\11\yolo11_PCFN.yaml.yaml")\
        .load(r'D:\bilibili\model\YOLO11\ultralytics-main\yolo11n.pt')  # build from YAML and transfer weights

    results = model.train(data=r'D:\bilibili\model\ultralytics-main\ultralytics\cfg\datasets\VOC_my.yaml',
                          epochs=100, imgsz=640, batch=8)

 

上一篇：DAY15|二叉树Part03|LeetCode: 222.完全二叉树的节点个数、110.平衡二叉树、257. 二叉树的所有路径、404.左叶子之和


下一篇：dedecms手机搜索不跳转手机页面模板的解决方法