本篇文章将介绍一个新的改进模块——SCConv（小波空间和通道重构卷积），并阐述如何将其应用于YOLOv11中，显著提升模型性能。为了减少YOLOv11模型的空间和通道维度上的冗余，我们引入空间和通道重构卷积。首先，我们将解析SCConv的工作原理，它通过空间重构单元（SRU）和通道重构单元（CRU）减少卷积神经网络中的空间和通道冗余。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，最终展现这一改进对目标检测效果的积极影响。
1. Spatial and Channel reconstruction Convolution(SCConv)结构介绍
SCConv模块由两个核心部分组成：空间重建单元 (SRU) 和 通道重建单元 (CRU)。它们按照顺序组合使用，首先通过SRU减少空间维度上的冗余，然后通过CRU减少通道维度上的冗余。SCConv可以无缝集成到现有的CNN中，用于替代标准卷积操作(Li_SCConv_Spatial_and_C…)。
1.1. 空间重建单元 (SRU)
SRU的主要目标是减少空间冗余。其工作流程如下：
-
分离操作：SRU通过训练好的参数对输入特征进行加权，分离出包含丰富空间信息的特征和不包含太多信息的冗余特征。
- 通过对特征图使用Group Normalization (GN)，提取每个特征图的缩放因子（即γ），γ反映了空间像素的方差，值越大，说明该特征图包含的空间信息越丰富。
- 基于这些缩放因子，SRU将特征图分为两部分：一部分包含丰富空间信息，另一部分则是较少的信息。
- 重建操作：SRU通过交叉重建的方式将分离的特征进行重组，提升信息流。该操作不仅减少了冗余，还通过将富有信息的特征与低信息的特征组合，进一步增强了特征的空间表达能力。
1.2. 通道重建单元 (CRU)
CRU的目标是减少通道维度上的冗余。其流程分为三个步骤：
-
分离 (Split)：CRU首先将输入特征图的通道分成两部分，分别包含αC个通道和(1-α)C个通道，然后通过1×1卷积进行压缩，减少计算量。
-
变换 (Transform)：分离后的上半部分特征图通过Group-wise Convolution (GWC) 和**Point-wise Convolution (PWC)**进行变换，提取代表性强的高层特征；下半部分通过廉价的1×1卷积提取浅层特征，作为补充。
-
融合 (Fuse)：通过全局平均池化（Pooling）和注意力机制，CRU将上半部分的高层特征和下半部分的浅层特征进行加权融合，得到最终的通道精炼特征。这种融合确保了信息在通道维度上的有效传递和冗余的消除。
2. YOLOv11与SCConv的结合
1. 改进C3k2：本文使用SCConv卷积改进C3k2，构建C3k2_SCConv模块，然后使用C3k2_SCConv替换原有的C3k2，这样就可以利用SCConv减少C3k2中的空间和通道的冗余。
2. 在backbone添加SCConv：本文将SCConv卷积添加到SPPF模块之前，减少backbone中的空间和通道的冗余。通过将空间和通道信息分别优化，减少冗余信息，从而提升模型的整体表现
3. Spatial and Channel reconstruction Convolution(SCConv)代码部分
import torch
import torch.nn.functional as F
import torch.nn as nn
from .conv import Conv
from .block import C2f, C3, Bottleneck
class GroupBatchnorm2d(nn.Module):
def __init__(self, c_num: int,
group_num: int = 16,
eps: float = 1e-10
):
super(GroupBatchnorm2d, self).__init__()
assert c_num >= group_num
self.group_num = group_num
self.weight = nn.Parameter(torch.randn(c_num, 1, 1))
self.bias = nn.Parameter(torch.zeros(c_num, 1, 1))
self.eps = eps
def forward(self, x):
N, C, H, W = x.size()
x = x.view(N, self.group_num, -1)
mean = x.mean(dim=2, keepdim=True)
std = x.std(dim=2, keepdim=True)
x = (x - mean) / (std + self.eps)
x = x.view(N, C, H, W)
return x * self.weight + self.bias
class SRU(nn.Module):
def __init__(self,
oup_channels: int,
group_num: int = 16,
gate_treshold: float = 0.5,
torch_gn: bool = False
):
super().__init__()
self.gn = nn.GroupNorm(num_channels=oup_channels, num_groups=group_num) if torch_gn else GroupBatchnorm2d(
c_num=oup_channels, group_num=group_num)
self.gate_treshold = gate_treshold
self.sigomid = nn.Sigmoid()
def forward(self, x):
gn_x = self.gn(x)
w_gamma = self.gn.weight / torch.sum(self.gn.weight)
w_gamma = w_gamma.view(1, -1, 1, 1)
reweigts = self.sigomid(gn_x * w_gamma)
# Gate
info_mask = reweigts >= self.gate_treshold
noninfo_mask = reweigts < self.gate_treshold
x_1 = info_mask * gn_x
x_2 = noninfo_mask * gn_x
x = self.reconstruct(x_1, x_2)
return x
def reconstruct(self, x_1, x_2):
x_11, x_12 = torch.split(x_1, x_1.size(1) // 2, dim=1)
x_21, x_22 = torch.split(x_2, x_2.size(1) // 2, dim=1)
return torch.cat([x_11 + x_22, x_12 + x_21], dim=1)
class CRU(nn.Module):
'''
alpha: 0<alpha<1
'''
def __init__(self,
op_channel: int,
alpha: float = 1 / 2,
squeeze_radio: int = 2,
group_size: int = 2,
group_kernel_size: int = 3,
):
super().__init__()
self.up_channel = up_channel = int(alpha * op_channel)
self.low_channel = low_channel = op_channel - up_channel
self.squeeze1 = nn.Conv2d(up_channel, up_channel // squeeze_radio, kernel_size=1, bias=False)
self.squeeze2 = nn.Conv2d(low_channel, low_channel // squeeze_radio, kernel_size=1, bias=False)
# up
self.GWC = nn.Conv2d(up_channel // squeeze_radio, op_channel, kernel_size=group_kernel_size, stride=1,
padding=group_kernel_size // 2, groups=group_size)
self.PWC1 = nn.Conv2d(up_channel // squeeze_radio, op_channel, kernel_size=1, bias=False)
# low
self.PWC2 = nn.Conv2d(low_channel // squeeze_radio, op_channel - low_channel // squeeze_radio, kernel_size=1,
bias=False)
self.advavg = nn.AdaptiveAvgPool2d(1)
def forward(self, x):
# Split
up, low = torch.split(x, [self.up_channel, self.low_channel], dim=1)
up, low = self.squeeze1(up), self.squeeze2(low)
# Transform
Y1 = self.GWC(up) + self.PWC1(up)
Y2 = torch.cat([self.PWC2(low), low], dim=1)
# Fuse
out = torch.cat([Y1, Y2], dim=1)
out = F.softmax(self.advavg(out), dim=1) * out
out1, out2 = torch.split(out, out.size(1) // 2, dim=1)
return out1 + out2
class ScConv(nn.Module):
def __init__(self,
op_channel: int,
group_num: int = 4,
gate_treshold: float = 0.5,
alpha: float = 1 / 2,
squeeze_radio: int = 2,
group_size: int = 2,
group_kernel_size: int = 3,
):
super().__init__()
self.SRU = SRU(op_channel,
group_num=group_num,
gate_treshold=gate_treshold)
self.CRU = CRU(op_channel,
alpha=alpha,
squeeze_radio=squeeze_radio,
group_size=group_size,
group_kernel_size=group_kernel_size)
def forward(self, x):
x = self.SRU(x)
x = self.CRU(x)
return x
class Bottleneck_ScConv(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 = ScConv(c2)
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_ScConv(c_, c_, shortcut, g, k=(k, k), e=1.0) for _ in range(n)))
# 在c3k=True时，使用Bottleneck_ScConv特征融合，为false的时候我们使用普通的Bottleneck提取特征
class C3k2_SC(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__':
DW = ScConv(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 =DW(input_tensor)
print("Input shape:",input_tensor.shape)
print("0utput shape:",output_tensor.shape)
4. 将SCConv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SCConv包
第三：在task.py中的模型配置部分下面代码
第一个改进需修改的地方
第二个改进，需修改的地方
elif m is ScConv:
args = [ch[f]]
第四：将模型配置文件复制到YOLOV11.YAMY文件中
第一个修改的配置文件
# 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_SC, [256, False, 0.25]]
- [-1, 1, Conv, [256, 3, 2]] # 3-P3/8
- [-1, 2, C3k2_SC, [512, False, 0.25]]
- [-1, 1, Conv, [512, 3, 2]] # 5-P4/16
- [-1, 2, C3k2_SC, [512, True]]
- [-1, 1, Conv, [1024, 3, 2]] # 7-P5/32
- [-1, 2, C3k2_SC, [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_SC, [512, False]] # 13
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 4], 1, Concat, [1]] # cat backbone P3
- [-1, 2, C3k2_SC, [256, False]] # 16 (P3/8-small)
- [-1, 1, Conv, [256, 3, 2]]
- [[-1, 13], 1, Concat, [1]] # cat head P4
- [-1, 2, C3k2_SC, [512, False]] # 19 (P4/16-medium)
- [-1, 1, Conv, [512, 3, 2]]
- [[-1, 10], 1, Concat, [1]] # cat head P5
- [-1, 2, C3k2_SC, [1024, True]] # 22 (P5/32-large)
- [[16, 19, 22], 1, Detect, [nc]] # Detect(P3, P4, P5)
第二个修改的配置文件
# 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, ScConv, []]
- [-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, [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, 14], 1, Concat, [1]] # cat head P4
- [-1, 2, C3k2, [512, False]] # 19 (P4/16-medium)
- [-1, 1, Conv, [512, 3, 2]]
- [[-1, 11], 1, Concat, [1]] # cat head P5
- [-1, 2, C3k2, [1024, True]] # 22 (P5/32-large)
- [[17, 20, 23], 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_SConv.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)
