改进YOLOv5系列：29.YOLOv5 结合 极简又强大的RepVGG 重参数化模型结构

最新创新点改进推荐

-💡统一使用 YOLO 代码框架，结合不同模块来构建不同的YOLO目标检测模型。

🔥 《芒果书》系列改进专栏内的改进文章，均包含多种模型改进方式，均适用于YOLOv3 、YOLOv4 、 YOLOR 、 YOLOX 、YOLOv5 、 YOLOv7 、 YOLOv8 改进（重点）！！！

🔥 专栏创新点教程 均有不少同学反应和我说已经在自己的数据集上有效涨点啦！！ 包括COCO数据集也能涨点

所有文章博客均包含 改进源代码部分，一键训练即可

🔥 对应专栏订阅的越早，就可以越早使用原创创新点去改进模型，抢先一步

以下《芒果书》改进YOLO专栏内容 适用于以下主流模型的改进
YOLOv3 、YOLOv4 、 YOLOR 、 YOLOX
YOLOv5 、 YOLOv7 、 YOLOv8
… 以上模型改进均适用《芒果书》🥭专栏系列

---

芒果书 点击以下链接 查看文章目录详情🔗

• 💡🎈☁️：一、CSDN原创《芒果改进YOLO高阶指南》强烈改进涨点推荐！📚推荐指数：🌟🌟🌟🌟🌟

• 💡🎈☁️：二、CSDN原创YOLO进阶 | 《芒果改进YOLO进阶指南》改进涨点推荐！📚推荐指数：🌟🌟🌟🌟🌟

• 💡🎈☁️：三、CSDN独家全网首发专栏 | 《目标检测YOLO改进指南》改进涨点推荐！推荐指数：🌟🌟🌟🌟🌟

---

本篇是《RepVGG结构🚀》的修改 演示

使用YOLOv5网络🚀作为示范，可以无缝加入到 YOLOv7、YOLOX、YOLOR、YOLOv4、Scaled_YOLOv4、YOLOv3等一系列YOLO算法模块

1.RepVGG模型理论部分

论文参考：最新RepVGG结构: Paper

模型定义

我们所说的“VGG式”指的是：

1. 没有任何分支结构。即通常所说的plain或feed-forward架构。

2. 仅使用3x3卷积。

3. 仅使用ReLU作为激活函数。

结构重参数化让VGG再次伟大

相比于各种多分支架构（如ResNet，Inception，DenseNet，各种NAS架构），近年来VGG式模型鲜有关注，主要自然是因为性能差。例如，有研究[1]认为，ResNet性能好的一种解释是ResNet的分支结构（shortcut）产生了一个大量子模型的隐式ensemble（因为每遇到一次分支，总的路径就变成两倍），单路架构显然不具备这种特点。

2.在YOLOv5中加入RepVGG模块🚀

使用YOLOv5算法🚀作为演示，模块可以无缝插入到YOLOv7、YOLOv5、YOLOv4、Scaled_YOLOv4、YOLOv3、YOLOR等一系列YOLO算法中

新增YOLOv5的yaml配置文件

首先增加以下yolov5_RepVGG.yaml文件,作为改进演示

# YOLOv5 🚀 by Ultralytics, GPL-3.0 license

# Parameters
nc: 80  # number of classes
depth_multiple: 0.33  # model depth multiple
width_multiple: 0.50  # 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 by yoloair
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, 1, RepVGGBlock, [128]], # 5-P4/16
   [-1, 1, Conv, [256, 3, 2]],  # 3-P3/8
   [-1, 6, RepVGGBlock, [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, SPPF, [1024, 5]],  # 9
  ]

# 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]],  # 13

   [-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]],  # 17 (P3/8-small)

   [-1, 1, Conv, [256, 3, 2]],
   [[-1, 14], 1, Concat, [1]],  # cat head P4
   [-1, 3, C3, [512, False]],  # 20 (P4/16-medium)

   [-1, 1, Conv, [512, 3, 2]],
   [[-1, 10], 1, Concat, [1]],  # cat head P5
   [-1, 3, C3, [1024, False]],  # 23 (P5/32-large)

   [[17, 20, 23], 1, Detect, [nc, anchors]],  # Detect(P3, P4, P5)
  ]

common.py配置

在./models/common.py文件中增加以下模块，直接复制即可

class RepVGGBlock(nn.Module):
    def __init__(self, in_channels, out_channels, kernel_size=3,
                 stride=1, padding=1, dilation=1, groups=1, padding_mode='zeros', deploy=False, use_se=False):
        super(RepVGGBlock, self).__init__()
        self.deploy = deploy
        self.groups = groups
        self.in_channels = in_channels
        padding_11 = padding - kernel_size // 2
        self.nonlinearity = nn.SiLU()
        # self.nonlinearity = nn.ReLU()
        if use_se:
            self.se = SEBlock(out_channels, internal_neurons=out_channels // 16)
        else:
            self.se = nn.Identity()
        if deploy:
            self.rbr_reparam = nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=kernel_size,
                                         stride=stride,
                                         padding=padding, dilation=dilation, groups=groups, bias=True,
                                         padding_mode=padding_mode)

        else:
            self.rbr_identity = nn.BatchNorm2d(
                num_features=in_channels) if out_channels == in_channels and stride == 1 else None
            self.rbr_dense = conv_bn(in_channels=in_channels, out_channels=out_channels, kernel_size=kernel_size,
                                     stride=stride, padding=padding, groups=groups)
            self.rbr_1x1 = conv_bn(in_channels=in_channels, out_channels=out_channels, kernel_size=1, stride=stride,
                                   padding=padding_11, groups=groups)
            # print('RepVGG Block, identity = ', self.rbr_identity)
def switch_to_deploy(self):
        if hasattr(self, 'rbr_1x1'):
            kernel, bias = self.get_equivalent_kernel_bias()
            self.rbr_reparam = nn.Conv2d(in_channels=self.rbr_dense.conv.in_channels, out_channels=self.rbr_dense.conv.out_channels,
                                    kernel_size=self.rbr_dense.conv.kernel_size, stride=self.rbr_dense.conv.stride,
                                    padding=self.rbr_dense.conv.padding, dilation=self.rbr_dense.conv.dilation, groups=self.rbr_dense.conv.groups, bias=True)
            self.rbr_reparam.weight.data = kernel
            self.rbr_reparam.bias.data = bias
            for para in self.parameters():
                para.detach_()
            self.rbr_dense = self.rbr_reparam
            # self.__delattr__('rbr_dense')
            self.__delattr__('rbr_1x1')
            if hasattr(self, 'rbr_identity'):
                self.__delattr__('rbr_identity')
            if hasattr(self, 'id_tensor'):
                self.__delattr__('id_tensor')
            self.deploy = True

    def get_equivalent_kernel_bias(self):
        kernel3x3, bias3x3 = self._fuse_bn_tensor(self.rbr_dense)
        kernel1x1, bias1x1 = self._fuse_bn_tensor(self.rbr_1x1)
        kernelid, biasid = self._fuse_bn_tensor(self.rbr_identity)
        return kernel3x3 + self._pad_1x1_to_3x3_tensor(kernel1x1) + kernelid, bias3x3 + bias1x1 + biasid

    def _pad_1x1_to_3x3_tensor(self, kernel1x1):
        if kernel1x1 is None:
            return 0
        else:
            return torch.nn.functional.pad(kernel1x1, [1, 1, 1, 1])

    def _fuse_bn_tensor(self, branch):
        if branch is None:
            return 0, 0
        if isinstance(branch, nn.Sequential):
            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
        else:
            assert isinstance(branch, nn.BatchNorm2d)
            if not hasattr(self, 'id_tensor'):
                input_dim = self.in_channels // self.groups
                kernel_value = np.zeros((self.in_channels, input_dim, 3, 3), dtype=np.float32)
                for i in range(self.in_channels):
                    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 forward(self, inputs):
        if self.deploy:
            return self.nonlinearity(self.rbr_dense(inputs))
        if hasattr(self, 'rbr_reparam'):
            return self.nonlinearity(self.se(self.rbr_reparam(inputs)))

        if self.rbr_identity is None:
            id_out = 0
        else:
            id_out = self.rbr_identity(inputs)

        return self.nonlinearity(self.se(self.rbr_dense(inputs) + self.rbr_1x1(inputs) + id_out))

yolo.py配置

然后找到./models/yolo.py文件下里的parse_model函数，将类名加入进去
在 models/yolo.py文件夹下

• parse_model函数中
• for i, (f, n, m, args) in enumerate(d['backbone'] + d['head']):内部
• 对应位置 下方只需要增加 RepVGGBlock模块

参考代码

elif m is RepVGGBlock:
            c1, c2 = ch[f], args[0]
            if c2 != no:  # if not output
                c2 = make_divisible(c2 * gw, 8)
            args = [c1, c2, *args[1:]]

训练yolov5_RepVGGBlock模型

python train.py --cfg yolov5_RepVGGBlock.yaml

推理过程效果

以下使用单独测试的RepVGG模块作为参考：

训练的时候代码
Model Summary: 375 layers, 5574845 parameters, 5574845 gradients, 16.2 GFLOPs

推理时候的代码
Model Summary: 284 layers, 5390365 parameters, 1567680 gradients, 15.7 GFLOPs

推理模型的数据相比于训练模型的数据

参数量、计算量、推理时间均有所减少

参考文献: 理论部分来自RepVGG作者的知乎文章：https://zhuanlan.zhihu.com/p/344324470