一、本文介绍

Hello，大家好，上一篇博客我们讲了利用 HGNetV2去替换YOLOv8的主干 ，经过结构的研究我们可以发现在HGNetV2的网络中有大量的卷积存在，所以我们可以用一种更加 轻量化 的卷积去优化HGNetV2从而达到更加轻量化的效果 （ 亲测优化后的HGNetV2网络比正常HGNetV2精度更高轻量化效果更好，非常适合轻量化的读者） ，同时HGNetV2的网络结构目前还没有推出论文，所以其理论知识在网络上也是非常的少，我也是根据网络结构图进行了分析，给大家进行讲解网络结构原理 （亲测替换之后主干GFLOPs降低到7.4，精度mAP提高0.06） 。

二、HGNetV2原理讲解

​

本文论文地址： RT-DETR论文地址

本文代码来源： HGNetV2的代码来源

​

2.1 HGNetV2的网络结构讲解

PP-HGNet 骨干网络的整体结构如下：

其中，PP-HGNet是由多个HG-Block组成，HG-Block的细节如下：

上面的图表是PP-HGNet 神经网络 架构的概览，下面我对其中的每一个模块进行分析：

1. Stem层： 这是网络的初始预处理层，通常包含 卷积层 ，开始从原始输入数据中提取特征。

2. HG（层次图）块： 这些块是网络的核心 组件 ，设计用于以层次化的方式处理数据。每个HG块可能处理数据的不同抽象层次，允许网络从低级和高级特征中学习。

3. LDS（可学习的下采样）层： 位于HG块之间的这些层可能执行下采样操作，减少特征图的空间维度，减少计算负载并可能增加后续层的感受野。

4. GAP（全局平均池化）： 在最终分类之前，使用GAP层将特征图的空间维度减少到每个特征图一个向量，有助于提高网络对输入数据空间变换的鲁棒性。

5. 最终的卷积和全连接（FC）层： 网络以一系列执行最终分类任务的层结束。这通常涉及一个卷积层（有时称为1x1卷积）来组合特征，然后是将这些特征映射到所需输出类别数量的全连接层。

这种架构的主要思想是利用层次化的方法来提取特征，其中复杂的模式可以在不同的规模和抽象层次上学习，提高网络处理复杂图像数据的能力。

这种分层和高效的处理对于图像分类等复杂任务非常有利，在这些任务中，精确预测至关重要的是在不同规模上识别复杂的模式和特征。图表还显示了HG块的扩展视图，包括多个不同滤波器大小的卷积层，以捕获多样化的特征，然后通过一个元素级相加或连接的操作（由+符号表示）在数据传递到下一层之前。

2.2 轻量化卷积

我这里利用的轻量化卷积只是官方仓库里面包含的四种，这个文章其实是给大家打开一个思路，这里的HGNet利用大量的卷积处理，所以我们能够替换其中大量的卷积从而达到优化和涨点的效果。

这几种卷积都是非常经典的了，其中RepConv只支持卷积核为3所以我也进行了一定的处理，原理就不再描述了。

三、HGNetV2的代码

这里的结构我们复制' ultralytics /nn/modules'目录下然后创建一个py文件粘贴进去即可，添加教程看章节四。

import torch
import torch.nn as nn
import math
import numpy as np
__all__ = ['Light_HGBlock']
def autopad(k, p=None, d=1):  # kernel, padding, dilation
    """Pad to 'same' shape outputs."""
    if d > 1:
        k = d * (k - 1) + 1 if isinstance(k, int) else [d * (x - 1) + 1 for x in k]  # actual kernel-size
    if p is None:
        p = k // 2 if isinstance(k, int) else [x // 2 for x in k]  # auto-pad
    return p
class Conv(nn.Module):
    """Standard convolution with args(ch_in, ch_out, kernel, stride, padding, groups, dilation, activation)."""
    default_act = nn.SiLU()  # default activation
    def __init__(self, c1, c2, k=1, s=1, p=None, g=1, d=1, act=True):
        """Initialize Conv layer with given arguments including activation."""
        super().__init__()
        self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p, d), groups=g, dilation=d, bias=False)
        self.bn = nn.BatchNorm2d(c2)
        self.act = self.default_act if act is True else act if isinstance(act, nn.Module) else nn.Identity()
    def forward(self, x):
        """Apply convolution, batch normalization and activation to input tensor."""
        return self.act(self.bn(self.conv(x)))
    def forward_fuse(self, x):
        """Perform transposed convolution of 2D data."""
        return self.act(self.conv(x))
class LightConv(nn.Module):
    """
    Light convolution with args(ch_in, ch_out, kernel).
    https://github.com/PaddlePaddle/PaddleDetection/blob/develop/ppdet/modeling/backbones/hgnet_v2.py
    """
    def __init__(self, c1, c2, k=1, act=nn.ReLU()):
        """Initialize Conv layer with given arguments including activation."""
        super().__init__()
        self.conv1 = Conv(c1, c2, 1, act=False)
        self.conv2 = DWConv(c2, c2, k, act=act)
    def forward(self, x):
        """Apply 2 convolutions to input tensor."""
        return self.conv2(self.conv1(x))
class DWConv(Conv):
    """Depth-wise convolution."""
    def __init__(self, c1, c2, k=1, s=1, d=1, act=True):  # ch_in, ch_out, kernel, stride, dilation, activation
        """Initialize Depth-wise convolution with given parameters."""
        super().__init__(c1, c2, k, s, g=math.gcd(c1, c2), d=d, act=act)
class DWConvTranspose2d(nn.ConvTranspose2d):
    """Depth-wise transpose convolution."""
    def __init__(self, c1, c2, k=1, s=1, p1=0, p2=0):  # ch_in, ch_out, kernel, stride, padding, padding_out
        """Initialize DWConvTranspose2d class with given parameters."""
        super().__init__(c1, c2, k, s, p1, p2, groups=math.gcd(c1, c2))
class ConvTranspose(nn.Module):
    """Convolution transpose 2d layer."""
    default_act = nn.SiLU()  # default activation
    def __init__(self, c1, c2, k=2, s=2, p=0, bn=True, act=True):
        """Initialize ConvTranspose2d layer with batch normalization and activation function."""
        super().__init__()
        self.conv_transpose = nn.ConvTranspose2d(c1, c2, k, s, p, bias=not bn)
        self.bn = nn.BatchNorm2d(c2) if bn else nn.Identity()
        self.act = self.default_act if act is True else act if isinstance(act, nn.Module) else nn.Identity()
    def forward(self, x):
        """Applies transposed convolutions, batch normalization and activation to input."""
        return self.act(self.bn(self.conv_transpose(x)))
    def forward_fuse(self, x):
        """Applies activation and convolution transpose operation to input."""
        return self.act(self.conv_transpose(x))
class Focus(nn.Module):
    """Focus wh information into c-space."""
    def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True):
        """Initializes Focus object with user defined channel, convolution, padding, group and activation values."""
        super().__init__()
        self.conv = Conv(c1 * 4, c2, k, s, p, g, act=act)
        # self.contract = Contract(gain=2)
    def forward(self, x):
        """
        Applies convolution to concatenated tensor and returns the output.
        Input shape is (b,c,w,h) and output shape is (b,4c,w/2,h/2).
        """
        return self.conv(torch.cat((x[..., ::2, ::2], x[..., 1::2, ::2], x[..., ::2, 1::2], x[..., 1::2, 1::2]), 1))
        # return self.conv(self.contract(x))
class GhostConv(nn.Module):
    """Ghost Convolution https://github.com/huawei-noah/ghostnet."""
    def __init__(self, c1, c2, k=1, s=1, g=1, act=True):
        """Initializes Ghost Convolution module with primary and cheap operations for efficient feature learning."""
        super().__init__()
        c_ = c2 // 2  # hidden channels
        self.cv1 = Conv(c1, c_, k, s, None, g, act=act)
        self.cv2 = Conv(c_, c_, 5, 1, None, c_, act=act)
    def forward(self, x):
        """Forward propagation through a Ghost Bottleneck layer with skip connection."""
        y = self.cv1(x)
        return torch.cat((y, self.cv2(y)), 1)
class RepConv(nn.Module):
    """
    RepConv is a basic rep-style block, including training and deploy status.
    This module is used in RT-DETR.
    Based on https://github.com/DingXiaoH/RepVGG/blob/main/repvgg.py
    """
    default_act = nn.SiLU()  # default activation
    def __init__(self, c1, c2, k=3, s=1, p=1, g=1, d=1, act=True, bn=False, deploy=False):
        """Initializes Light Convolution layer with inputs, outputs & optional activation function."""
        super().__init__()
        assert k == 3 and p == 1
        self.g = g
        self.c1 = c1
        self.c2 = c2
        self.act = self.default_act if act is True else act if isinstance(act, nn.Module) else nn.Identity()
        self.bn = nn.BatchNorm2d(num_features=c1) if bn and c2 == c1 and s == 1 else None
        self.conv1 = Conv(c1, c2, k, s, p=p, g=g, act=False)
        self.conv2 = Conv(c1, c2, 1, s, p=(p - k // 2), g=g, act=False)
    def forward_fuse(self, x):
        """Forward process."""
        return self.act(self.conv(x))
    def forward(self, x):
        """Forward process."""
        id_out = 0 if self.bn is None else self.bn(x)
        return self.act(self.conv1(x) + self.conv2(x) + id_out)
    def get_equivalent_kernel_bias(self):
        """Returns equivalent kernel and bias by adding 3x3 kernel, 1x1 kernel and identity kernel with their biases."""
        kernel3x3, bias3x3 = self._fuse_bn_tensor(self.conv1)
        kernel1x1, bias1x1 = self._fuse_bn_tensor(self.conv2)
        kernelid, biasid = self._fuse_bn_tensor(self.bn)
        return kernel3x3 + self._pad_1x1_to_3x3_tensor(kernel1x1) + kernelid, bias3x3 + bias1x1 + biasid
    @staticmethod
    def _pad_1x1_to_3x3_tensor(kernel1x1):
        """Pads a 1x1 tensor to a 3x3 tensor."""
        if kernel1x1 is None:
            return 0
        else:
            return torch.nn.functional.pad(kernel1x1, [1, 1, 1, 1])
    def _fuse_bn_tensor(self, branch):
        """Generates appropriate kernels and biases for convolution by fusing branches of the neural network."""
        if branch is None:
            return 0, 0
        if isinstance(branch, Conv):
            kernel = branch.conv.weight
            running_mean = branch.bn.running_mean
            running_var = branch.bn.running_var
            gamma = branch.bn.weight
            beta = branch.bn.bias
            eps = branch.bn.eps
        elif isinstance(branch, nn.BatchNorm2d):
            if not hasattr(self, "id_tensor"):
                input_dim = self.c1 // self.g
                kernel_value = np.zeros((self.c1, input_dim, 3, 3), dtype=np.float32)
                for i in range(self.c1):
                    kernel_value[i, i % input_dim, 1, 1] = 1
                self.id_tensor = torch.from_numpy(kernel_value).to(branch.weight.device)
            kernel = self.id_tensor
            running_mean = branch.running_mean
            running_var = branch.running_var
            gamma = branch.weight
            beta = branch.bias
            eps = branch.eps
        std = (running_var + eps).sqrt()
        t = (gamma / std).reshape(-1, 1, 1, 1)
        return kernel * t, beta - running_mean * gamma / std
    def fuse_convs(self):
        """Combines two convolution layers into a single layer and removes unused attributes from the class."""
        if hasattr(self, "conv"):
            return
        kernel, bias = self.get_equivalent_kernel_bias()
        self.conv = nn.Conv2d(
            in_channels=self.conv1.conv.in_channels,
            out_channels=self.conv1.conv.out_channels,
            kernel_size=self.conv1.conv.kernel_size,
            stride=self.conv1.conv.stride,
            padding=self.conv1.conv.padding,
            dilation=self.conv1.conv.dilation,
            groups=self.conv1.conv.groups,
            bias=True,
        ).requires_grad_(False)
        self.conv.weight.data = kernel
        self.conv.bias.data = bias
        for para in self.parameters():
            para.detach_()
        self.__delattr__("conv1")
        self.__delattr__("conv2")
        if hasattr(self, "nm"):
            self.__delattr__("nm")
        if hasattr(self, "bn"):
            self.__delattr__("bn")
        if hasattr(self, "id_tensor"):
            self.__delattr__("id_tensor")
class Light_HGBlock(nn.Module):
    """
    HG_Block of PPHGNetV2 with 2 convolutions and LightConv.
    https://github.com/PaddlePaddle/PaddleDetection/blob/develop/ppdet/modeling/backbones/hgnet_v2.py
    """
    def __init__(self, c1, cm, c2, k=3, n=6, num=1, shortcut=False, act=True):
        """Initializes a CSP Bottleneck with 1 convolution using specified input and output channels."""
        super().__init__()
        block = Conv
        if num == 1:
            block = GhostConv
        elif num == 2:
            block = RepConv  # RepConv Only supported k = 3
            k = 3
        elif num == 3:
            block = DWConv
        elif num == 4:
            block = LightConv
        self.m = nn.ModuleList(block(c1 if i == 0 else cm, cm, k=k, act=act) for i in range(n))
        self.sc = Conv(c1 + n * cm, c2 // 2, 1, 1, act=act)  # squeeze conv
        self.ec = Conv(c2 // 2, c2, 1, 1, act=act)  # excitation conv
        self.add = shortcut and c1 == c2
    def forward(self, x):
        """Forward pass of a PPHGNetV2 backbone layer."""
        y = [x]
        y.extend(m(y[-1]) for m in self.m)
        y = self.ec(self.sc(torch.cat(y, 1)))
        return y + x if self.add else y

四、手把手教你添加HGNetV2

4.1 手把手教你添加HGNetV2

这里的添加方法比较特殊，和之前的都不一样，但是也很简单修改三处即可使用。

4.1.1 修改一

我们找到如下的文件'ultralytics/nn/tasks.py'，在开头的地方导入我们刚才创建文件的模块。

4.1.2 修改二

按照如下图所示添加即可

4.1.3 修改三

本文以及默认大家用的是以及集成过RT-DETR代码的ultralytics仓库了（其中以及包含了HGNet的代码文件），所以我们只需要添加几行代码就能够回复官方删除掉的功能。

我们首先需要找到'ultralytics/nn/tasks.py'文件然后找到'def parse_model(d, ch, verbose=True):  # model_dict, input_channels(3)'

下面的代码我们看到大概基本的样子的.

复制此处的代码按照下面的图片进行添加即可，不要自己打！

            cm = make_divisible(min(cm, max_channels) * width, 8)
            c2 = make_divisible(min(c2, max_channels) * width, 8)
            n = n_ = max(round(n * depth), 1) if n > 1 else n  # depth gain

到此我们就完成了官方的代码修复，我们此时运行代码比如V8n那么你就会发现的大幅度的减少了参数量，

4. 2 使用说明

这里有一个参数，需要大家修改，在章节2.2我说了我用了好几个卷积可以提供大家选择，所以这里分别可以有四个卷积可以给大家使用，大家可以看下面的代码。

        if num == 1:
            block = GhostConv
        elif num == 2:
            block = RepConv  # RepConv Only supported k = 3
            k = 3
        elif num == 3:
            block = DWConv
        elif num == 4:
            block = LightConv

如果我们想要使用RepConv为例，那么我们修改图中的红框位置的地方我们设置为2此时使用的就是RepConv，默认使用的是3也就是DWConv，实验结果也是这个跑出来的。

4.3 HGNet-l的yaml文件

此版本训练信息：YOLO11-LightHGNetV2-l summary: 424 layers, 2,381,262 parameters, 2,381,246 gradients, 7.1 GFLOPs

# 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, HGStem, [32, 48]]  # 0-P2/4
  - [-1, 6, Light_HGBlock, [48, 128, 3]]  # stage 1
  - [-1, 1, DWConv, [128, 3, 2, 1, 1]]  # 2-P3/8
  - [-1, 6, Light_HGBlock, [96, 512, 3]]   # stage 2
  - [-1, 1, DWConv, [512, 3, 2, 2, False]]  # 4-P3/16
  - [-1, 6, Light_HGBlock, [192, 1024, 5, 3, False]]  # cm, c2, k, light, shortcut
  - [-1, 6, Light_HGBlock, [192, 1024, 5, 3, True]]
  - [-1, 6, Light_HGBlock, [192, 1024, 5, 3, True]]  # stage 3
  - [-1, 1, DWConv, [1024, 3, 2, 1, False]]  # 8-P4/32
  - [-1, 6, Light_HGBlock, [384, 2048, 5, 3, False]]  # stage 4
  - [-1, 1, SPPF, [1024, 5]] # 10
  - [-1, 1, C2PSA, [1024]] # 11
# YOLOv10.0n head
head:
  - [-1, 1, nn.Upsample, [None, 2, "nearest"]] # 12
  - [[-1, 7], 1, Concat, [1]] # cat backbone P4
  - [-1, 2, C3k2, [512, False]] # 14
  - [-1, 1, nn.Upsample, [None, 2, "nearest"]]
  - [[-1, 3], 1, Concat, [1]] # cat backbone P3
  - [-1, 2, C3k2, [256, False]] # 17 (P3/8-small)
  - [-1, 1, Conv, [256, 3, 2]]
  - [[-1, 14], 1, Concat, [1]] # cat head P4
  - [-1, 2, C3k2, [512, False]] # 20 (P4/16-medium)
  - [-1, 1, SCDown, [512, 3, 2]]
  - [[-1, 11], 1, Concat, [1]] # cat head P5
  - [-1, 2, C3k2, [1024, True]] # 23 (P5/32-large)
  - [[17, 20, 23], 1, v10Detect, [nc]] # Detect(P3, P4, P5)

4.4 HGNetV2-x的yaml文件

此版本训练信息：YOLO11-LightHGNetV2-x summary: 460 layers, 2,643,782 parameters, 2,643,766 gradients, 8.0 GFLOPs

# 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, HGStem, [32, 64]]  # 0-P2/4
  - [-1, 6, Light_HGBlock, [64, 128, 3]]  # stage 1
  - [-1, 1, DWConv, [128, 3, 2, 1, False]]  # 2-P3/8
  - [-1, 6, Light_HGBlock, [128, 512, 3]]
  - [-1, 6, Light_HGBlock, [128, 512, 3, 3, True]]   # 4-stage 2
  - [-1, 1, DWConv, [512, 3, 2, 1, False]]  # 5-P3/16
  - [-1, 6, Light_HGBlock, [256, 1024, 5, 3, False]]  # cm, c2, k, light, shortcut
  - [-1, 6, Light_HGBlock, [256, 1024, 5, 3, True]]
  - [-1, 6, Light_HGBlock, [256, 1024, 5, 3, True]]
  - [-1, 6, Light_HGBlock, [256, 1024, 5, 3, True]]
  - [-1, 6, Light_HGBlock, [256, 1024, 5, 3, True]]  # 10-stage 3
  - [-1, 1, DWConv, [1024, 3, 2, 1, False]]  # 11-P4/32
  - [-1, 6, Light_HGBlock, [512, 2048, 5, 3, False]]
  - [-1, 6, Light_HGBlock, [512, 2048, 5, 3, True]]  # 13-stage 4
  - [-1, 1, SPPF, [1024, 5]] # 14
  - [-1, 1, PSA, [1024]] # 15
# YOLOv10.0n head
head:
  - [-1, 1, nn.Upsample, [None, 2, "nearest"]]
  - [[-1, 10], 1, Concat, [1]] # cat backbone P4
  - [-1, 2, C3k2, [512, False]] # 18
  - [-1, 1, nn.Upsample, [None, 2, "nearest"]]
  - [[-1, 4], 1, Concat, [1]] # cat backbone P3
  - [-1, 2, C3k2, [256, False]] # 21 (P3/8-small)
  - [-1, 1, Conv, [256, 3, 2]]
  - [[-1, 18], 1, Concat, [1]] # cat head P4
  - [-1, 2, C3k2, [512, False]] # 24 (P4/16-medium)
  - [-1, 1, SCDown, [512, 3, 2]]
  - [[-1, 15], 1, Concat, [1]] # cat head P5
  - [-1, 2, C3k2, [1024, False]] # 27 (P5/32-large)
  - [[21, 24, 27], 1, v10Detect, [nc]] # Detect(P3, P4, P5)

4.5 训练代码

import warnings
warnings.filterwarnings('ignore')
from ultralytics import YOLO
if __name__ == '__main__':
    model = YOLO('ultralytics/cfg/models/v8/yolov8-C2f-FasterBlock.yaml')
    # model.load('yolov8n.pt') # loading pretrain weights
    model.train(data=r'替换数据集yaml文件地址',
                # 如果大家任务是其它的'ultralytics/cfg/default.yaml'找到这里修改task可以改成detect, segment, classify, pose
                cache=False,
                imgsz=640,
                epochs=150,
                single_cls=False,  # 是否是单类别检测
                batch=4,
                close_mosaic=10,
                workers=0,
                device='0',
                optimizer='SGD', # using SGD
                # resume='', # 如过想续训就设置last.pt的地址
                amp=False,  # 如果出现训练损失为Nan可以关闭amp
                project='runs/train',
                name='exp',
                )

五、运行成功记录

下面的图片是证明成功运行的截图，确保我发的改进机制是可用的。

六、本文总结

到此本文的正式分享内容就结束了，在这里给大家推荐我的YOLOv11改进有效涨点专栏，本专栏目前为新开的平均质量分98分，后期我会根据各种最新的前沿顶会进行论文复现，也会对一些老的改进机制进行补充， 目前本专栏免费阅读(暂时，大家尽早关注不迷路~) ，如果大家觉得本文帮助到你了，订阅本专栏，关注后续更多的更新~

​