【YOLOv12多模态融合改进】| CGA Fusion：内容引导的注意力融合模块，空间权重引导的多模态特征自适应融合
一、本文介绍
本文记录的是利用 CGA Fusion 模块改进 YOLOv12 的多模态融合部分 。
CGAFusion
（Content-Guided Attention Fusion）通过
内容引导注意力生成空间权重
，
引导高低层特征的自适应融合
。本文利用
CGA Fusion
模块，通过内容引
导注意力生成空间权重，自适应地融合两个模态的特征
，
在特征融合阶段实现跨模态语义对齐与噪声抑制
，增强对不同模态互补特征的利用能力，从而提升模型在多模态场景下的检测鲁棒性与准确性。
二、CGA Fusion模块介绍
DEA-Net: Single image dehazing based on detail-enhanced convolution and content-guided attention
2.1 CGA模块（Content-Guided Attention）
2.1.1 设计出发点
传统注意力机制（如FAM、CBAM）存在两大缺陷：
- 无法处理特征级雾霾不均匀性 ：现有方法仅关注图像级雾霾分布（如空间注意力），忽略特征级通道间的雾霾差异。不同通道的特征编码不同语义（如边缘、纹理），需独立分配空间重要性图（SIM）。
- 通道与空间注意力缺乏交互 ：传统模块（如FAM）顺序计算通道和空间注意力，未融合两者信息，导致特征校准不全面。
目标 ：生成 通道特定的空间重要性图（Channel-specific SIM） ，同时融合通道与空间注意力，提升特征表示能力。
2.1.2 原理与结构
CGA采用 粗到细（Coarse-to-Fine） 的两阶段注意力生成机制，结构如图所示：
-
阶段1：生成粗粒度SIM（ W c o a W_{coa} W co a ）
- 通道注意力（ W c W_c W c ） ：通过全局平均池化（GAP）压缩空间维度，经两层1×1卷积生成通道权重，强调关键通道。
- 空间注意力（ W s W_s W s ） ：对通道维度进行GAP和全局最大池化（GMP），经7×7卷积生成空间重要性图，捕捉图像级雾霾分布。
- 融合粗粒度SIM ：通过元素相加融合 W c W_c W c 和 W s W_s W s ，得到初步的通道共享空间注意力图 W c o a W_{coa} W co a 。
-
阶段2：细化为通道特定SIM（(W)）
- 通道洗牌（Channel Shuffle） ：将输入特征 X X X 与 W c o a W_{coa} W co a 按通道交替排列，促进跨通道信息交互。
- 组卷积（Group Convolution） ：通过7×7组卷积（组数等于通道数），利用输入特征内容引导每个通道的SIM细化，生成最终通道特定的 W W W 。
公式总结
：
W
c
o
a
=
W
c
+
W
s
,
W
=
σ
(
G
C
7
×
7
(
C
S
(
[
X
,
W
c
o
a
]
)
)
)
,
\begin{aligned} W_{coa} &= W_c + W_s, \\ W &= \sigma\left(\mathcal{G}C_{7×7}\left(CS\left([X, W_{coa}]\right)\right)\right), \end{aligned}
W
co
a
W
=
W
c
+
W
s
,
=
σ
(
G
C
7
×
7
(
CS
(
[
X
,
W
co
a
]
)
)
)
,
其中
σ
\sigma
σ
为Sigmoid激活函数，
C
S
(
⋅
)
CS(\cdot)
CS
(
⋅
)
为通道洗牌，
G
C
\mathcal{G}C
G
C
为组卷积。
2.1.3 优势
- 通道特异性 ：为每个通道分配唯一SIM，精准捕捉不同通道的雾霾分布差异（如纹理丰富区域 vs. 平滑区域），避免传统单通道SIM的语义混淆。
- 信息交互 ：融合通道与空间注意力，通过内容引导（输入特征 X X X ）动态调整注意力权重，提升特征表达的灵活性。
2.2 CGA Fusion模块（CGA-Based Mixup Fusion Scheme）
2.2.1 设计出发点
传统特征融合（如相加、拼接）存在 感受野不匹配问题 ：
- 浅层特征（低层次，如边缘）与深层特征（高层次，如语义）的感受野差异显著，简单融合无法有效对齐信息。
- 现有混合融合（Mixup）依赖固定权重或自学习标量权重，未考虑空间位置差异。
目标 ：利用CGA生成的空间权重，自适应融合高低层特征，解决感受野不匹配并增强梯度流通。
2.2.2 原理与结构
模块结构如图所示：
- 输入 ：编码器的低层特征（ F l o w F_{low} F l o w ）和解码器的高层特征（ F h i g h F_{high} F hi g h ）。
- 注意力权重生成 ：将 F l o w F_{low} F l o w 和 F h i g h F_{high} F hi g h 输入CGA，生成空间权重 W W W （范围0-1），其中每个像素值表示低层特征的融合比例。
-
加权融合
：
F fuse = C 1 × 1 ( F l o w ⋅ W + F h i g h ⋅ ( 1 − W ) + F l o w + F h i g h ) , F_{\text{fuse}} = \mathcal{C}_{1×1}\left(F_{low} \cdot W + F_{high} \cdot (1-W) + F_{low} + F_{high}\right), F fuse = C 1 × 1 ( F l o w ⋅ W + F hi g h ⋅ ( 1 − W ) + F l o w + F hi g h ) ,
其中 C 1 × 1 \mathcal{C}_{1×1} C 1 × 1 为1×1卷积，用于特征投影。通过 W W W 动态调节高低层特征的贡献，并通过跳跃连接（ F l o w + F h i g h F_{low} + F_{high} F l o w + F hi g h ）缓解梯度消失。
2.2.3 优势
- 自适应融合 ：空间权重 W W W 根据特征内容动态调整，例如在边缘区域增强低层特征，在语义区域侧重高层特征，解决感受野不匹配。
- 梯度优化 ：跳跃连接和加权求和机制增强了浅层到深层的信息流动，提升训练稳定性，PSNR较传统Mixup提升0.23 dB。
论文： https://export.arxiv.org/pdf/2301.04805
源码： https://github.com/cecret3350/DEA-Net
三、CGA Fusion的实现代码
CGA Fusion
的实现代码如下：
import torch.nn as nn
import torch
from einops import rearrange
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 SpatialAttention_CGA(nn.Module):
def __init__(self):
super(SpatialAttention_CGA, self).__init__()
self.sa = nn.Conv2d(2, 1, 7, padding=3, padding_mode='reflect' ,bias=True)
def forward(self, x):
x_avg = torch.mean(x, dim=1, keepdim=True)
x_max, _ = torch.max(x, dim=1, keepdim=True)
x2 = torch.concat([x_avg, x_max], dim=1)
sattn = self.sa(x2)
return sattn
class ChannelAttention_CGA(nn.Module):
def __init__(self, dim, reduction = 8):
super(ChannelAttention_CGA, self).__init__()
self.gap = nn.AdaptiveAvgPool2d(1)
self.ca = nn.Sequential(
nn.Conv2d(dim, dim // reduction, 1, padding=0, bias=True),
nn.ReLU(inplace=True),
nn.Conv2d(dim // reduction, dim, 1, padding=0, bias=True),
)
def forward(self, x):
x_gap = self.gap(x)
cattn = self.ca(x_gap)
return cattn
class PixelAttention_CGA(nn.Module):
def __init__(self, dim):
super(PixelAttention_CGA, self).__init__()
self.pa2 = nn.Conv2d(2 * dim, dim, 7, padding=3, padding_mode='reflect' ,groups=dim, bias=True)
self.sigmoid = nn.Sigmoid()
def forward(self, x, pattn1):
B, C, H, W = x.shape
x = x.unsqueeze(dim=2) # B, C, 1, H, W
pattn1 = pattn1.unsqueeze(dim=2) # B, C, 1, H, W
x2 = torch.cat([x, pattn1], dim=2) # B, C, 2, H, W
x2 = rearrange(x2, 'b c t h w -> b (c t) h w')
pattn2 = self.pa2(x2)
pattn2 = self.sigmoid(pattn2)
return pattn2
class CGAFusion(nn.Module):
def __init__(self, dim, reduction=8):
super(CGAFusion, self).__init__()
self.sa = SpatialAttention_CGA()
self.ca = ChannelAttention_CGA(dim, reduction)
self.pa = PixelAttention_CGA(dim)
self.conv = nn.Conv2d(dim, dim, 1, bias=True)
self.sigmoid = nn.Sigmoid()
def forward(self, data):
x, y = data
initial = x + y
cattn = self.ca(initial)
sattn = self.sa(initial)
pattn1 = sattn + cattn
pattn2 = self.sigmoid(self.pa(initial, pattn1))
result = initial + pattn2 * x + (1 - pattn2) * y
result = self.conv(result)
return result
四、融合步骤
5.1 修改一
① 在
ultralytics/nn/
目录下新建
AddModules
文件夹用于存放模块代码
② 在
AddModules
文件夹下新建
CGAFusion .py
，将
第三节
中的代码粘贴到此处
5.2 修改二
在
AddModules
文件夹下新建
__init__.py
（已有则不用新建），在文件内导入模块：
from .CGAFusion import *
5.3 修改三
在
ultralytics/nn/modules/tasks.py
文件中，需要在两处位置添加各模块类名称。
首先：导入模块
其次：在
parse_model函数
中注册
CGAFusion
模块
elif m in {CGAFusion}:
c2 = ch[f[0]]
args = [c2]
最后将
ultralytics/utils/torch_utils.py
中的
get_flops
函数中的
stride
指定为
640
。
五、yaml模型文件
5.1 中期融合⭐
📌 此模型的修方法是将原本的中期融合中的Concat融合部分换成CGAFusion，融合骨干部分的多模态信息。
# YOLOv12 🚀, AGPL-3.0 license
# YOLOv12 object detection model with P3-P5 outputs. For Usage examples see https://docs.ultralytics.com/tasks/detect
# CFG file for YOLOv12-turbo
# Parameters
ch: 6
nc: 80 # number of classes
scales: # model compound scaling constants, i.e. 'model=yolov12n.yaml' will call yolov12.yaml with scale 'n'
# [depth, width, max_channels]
n: [0.50, 0.25, 1024] # summary: 497 layers, 2,553,904 parameters, 2,553,888 gradients, 6.2 GFLOPs
s: [0.50, 0.50, 1024] # summary: 497 layers, 9,127,424 parameters, 9,127,408 gradients, 19.7 GFLOPs
m: [0.50, 1.00, 512] # summary: 533 layers, 19,670,784 parameters, 19,670,768 gradients, 60.4 GFLOPs
l: [1.00, 1.00, 512] # summary: 895 layers, 26,506,496 parameters, 26,506,480 gradients, 83.3 GFLOPs
x: [1.00, 1.50, 512] # summary: 895 layers, 59,414,176 parameters, 59,414,160 gradients, 185.9 GFLOPs
# YOLO12 backbone
backbone:
# [from, repeats, module, args]
- [-1, 1, IN, []] # 0
- [-1, 1, Multiin, [1]] # 1
- [-2, 1, Multiin, [2]] # 2
- [1, 1, Conv, [64, 3, 2]] # 3-P1/2
- [-1, 1, Conv, [128, 3, 2, 1, 2]] # 4-P2/4
- [-1, 2, C3k2, [256, False, 0.25]]
- [-1, 1, Conv, [256, 3, 2, 1, 4]] # 6-P3/8
- [-1, 2, C3k2, [512, False, 0.25]]
- [-1, 1, Conv, [512, 3, 2]] # 8-P4/16
- [-1, 4, A2C2f, [512, True, 4]]
- [-1, 1, Conv, [1024, 3, 2]] # 10-P5/32
- [-1, 4, A2C2f, [1024, True, 1]] # 11
- [2, 1, Conv, [64, 3, 2]] # 12-P1/2
- [-1, 1, Conv, [128, 3, 2, 1, 2]] # 13-P2/4
- [-1, 2, C3k2, [256, False, 0.25]]
- [-1, 1, Conv, [256, 3, 2, 1, 4]] # 15-P3/8
- [-1, 2, C3k2, [512, False, 0.25]]
- [-1, 1, Conv, [512, 3, 2]] # 17-P4/16
- [-1, 4, A2C2f, [512, True, 4]]
- [-1, 1, Conv, [1024, 3, 2]] # 19-P5/32
- [-1, 4, A2C2f, [1024, True, 1]] # 20
- [[7, 16], 1, CGAFusion, []] # 21 cat backbone P3
- [[9, 18], 1, CGAFusion, []] # 22 cat backbone P4
- [[11, 20], 1, CGAFusion, []] # 23 cat backbone P5
# YOLO12 head
head:
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 22], 1, Concat, [1]] # cat backbone P4
- [-1, 2, A2C2f, [512, False, -1]] # 26
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 21], 1, Concat, [1]] # cat backbone P3
- [-1, 2, A2C2f, [256, False, -1]] # 29
- [-1, 1, Conv, [256, 3, 2]]
- [[-1, 26], 1, Concat, [1]] # cat head P4
- [-1, 2, A2C2f, [512, False, -1]] # 32
- [-1, 1, Conv, [512, 3, 2]]
- [[-1, 23], 1, Concat, [1]] # cat head P5
- [-1, 2, C3k2, [1024, True]] # 35 (P5/32-large)
- [[29, 32, 35], 1, Detect, [nc]] # Detect(P3, P4, P5)
5.2 中-后期融合⭐
📌 此模型的修方法是将原本的中-后期融合中的Concat融合部分换成CGAFusion，融合FPN部分的多模态信息。
# YOLOv12 🚀, AGPL-3.0 license
# YOLOv12 object detection model with P3-P5 outputs. For Usage examples see https://docs.ultralytics.com/tasks/detect
# CFG file for YOLOv12-turbo
# Parameters
ch: 6
nc: 80 # number of classes
scales: # model compound scaling constants, i.e. 'model=yolov12n.yaml' will call yolov12.yaml with scale 'n'
# [depth, width, max_channels]
n: [0.50, 0.25, 1024] # summary: 497 layers, 2,553,904 parameters, 2,553,888 gradients, 6.2 GFLOPs
s: [0.50, 0.50, 1024] # summary: 497 layers, 9,127,424 parameters, 9,127,408 gradients, 19.7 GFLOPs
m: [0.50, 1.00, 512] # summary: 533 layers, 19,670,784 parameters, 19,670,768 gradients, 60.4 GFLOPs
l: [1.00, 1.00, 512] # summary: 895 layers, 26,506,496 parameters, 26,506,480 gradients, 83.3 GFLOPs
x: [1.00, 1.50, 512] # summary: 895 layers, 59,414,176 parameters, 59,414,160 gradients, 185.9 GFLOPs
# YOLO12 backbone
backbone:
# [from, repeats, module, args]
- [-1, 1, IN, []] # 0
- [-1, 1, Multiin, [1]] # 1
- [-2, 1, Multiin, [2]] # 2
- [1, 1, Conv, [64, 3, 2]] # 3-P1/2
- [-1, 1, Conv, [128, 3, 2, 1, 2]] # 4-P2/4
- [-1, 2, C3k2, [256, False, 0.25]]
- [-1, 1, Conv, [256, 3, 2, 1, 4]] # 6-P3/8
- [-1, 2, C3k2, [512, False, 0.25]]
- [-1, 1, Conv, [512, 3, 2]] # 8-P4/16
- [-1, 4, A2C2f, [512, True, 4]]
- [-1, 1, Conv, [1024, 3, 2]] # 10-P5/32
- [-1, 4, A2C2f, [1024, True, 1]] # 11
- [2, 1, Conv, [64, 3, 2]] # 12-P1/2
- [-1, 1, Conv, [128, 3, 2, 1, 2]] # 13-P2/4
- [-1, 2, C3k2, [256, False, 0.25]]
- [-1, 1, Conv, [256, 3, 2, 1, 4]] # 15-P3/8
- [-1, 2, C3k2, [512, False, 0.25]]
- [-1, 1, Conv, [512, 3, 2]] # 17-P4/16
- [-1, 4, A2C2f, [512, True, 4]]
- [-1, 1, Conv, [1024, 3, 2]] # 19-P5/32
- [-1, 4, A2C2f, [1024, True, 1]] # 20
# YOLO12 head
head:
- [11, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 9], 1, Concat, [1]] # cat backbone P4
- [-1, 2, A2C2f, [512, False, -1]] # 23
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 7], 1, Concat, [1]] # cat backbone P3
- [-1, 2, A2C2f, [256, False, -1]] # 26
- [20, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 18], 1, Concat, [1]] # cat backbone P4
- [-1, 2, A2C2f, [512, False, -1]] # 29
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 16], 1, Concat, [1]] # cat backbone P3
- [-1, 2, A2C2f, [256, False, -1]] # 32
- [[11, 20], 1, CGAFusion, []] # cat head P5 33
- [[23, 29], 1, CGAFusion, []] # cat head P5 34
- [[26, 32], 1, CGAFusion, []] # cat head P5 35
- [-1, 1, Conv, [256, 3, 2]]
- [[-1, 34], 1, Concat, [1]] # cat head P4
- [-1, 2, A2C2f, [512, False, -1]] # 38
- [-1, 1, Conv, [512, 3, 2]]
- [[-1, 33], 1, Concat, [1]] # cat head P5
- [-1, 2, C3k2, [1024, True]] # 41 (P5/32-large)
- [[35, 38, 41], 1, Detect, [nc]] # Detect(P3, P4, P5)
5.3 后期融合⭐
📌 此模型的修方法是将原本的后期融合中的Concat融合部分换成CGAFusion，融合颈部部分的多模态信息。
# YOLOv12 🚀, AGPL-3.0 license
# YOLOv12 object detection model with P3-P5 outputs. For Usage examples see https://docs.ultralytics.com/tasks/detect
# CFG file for YOLOv12-turbo
# Parameters
ch: 6
nc: 80 # number of classes
scales: # model compound scaling constants, i.e. 'model=yolov12n.yaml' will call yolov12.yaml with scale 'n'
# [depth, width, max_channels]
n: [0.50, 0.25, 1024] # summary: 497 layers, 2,553,904 parameters, 2,553,888 gradients, 6.2 GFLOPs
s: [0.50, 0.50, 1024] # summary: 497 layers, 9,127,424 parameters, 9,127,408 gradients, 19.7 GFLOPs
m: [0.50, 1.00, 512] # summary: 533 layers, 19,670,784 parameters, 19,670,768 gradients, 60.4 GFLOPs
l: [1.00, 1.00, 512] # summary: 895 layers, 26,506,496 parameters, 26,506,480 gradients, 83.3 GFLOPs
x: [1.00, 1.50, 512] # summary: 895 layers, 59,414,176 parameters, 59,414,160 gradients, 185.9 GFLOPs
# YOLO12 backbone
backbone:
# [from, repeats, module, args]
- [-1, 1, IN, []] # 0
- [-1, 1, Multiin, [1]] # 1
- [-2, 1, Multiin, [2]] # 2
- [1, 1, Conv, [64, 3, 2]] # 3-P1/2
- [-1, 1, Conv, [128, 3, 2, 1, 2]] # 4-P2/4
- [-1, 2, C3k2, [256, False, 0.25]]
- [-1, 1, Conv, [256, 3, 2, 1, 4]] # 6-P3/8
- [-1, 2, C3k2, [512, False, 0.25]]
- [-1, 1, Conv, [512, 3, 2]] # 8-P4/16
- [-1, 4, A2C2f, [512, True, 4]]
- [-1, 1, Conv, [1024, 3, 2]] # 10-P5/32
- [-1, 4, A2C2f, [1024, True, 1]] # 11
- [2, 1, Conv, [64, 3, 2]] # 12-P1/2
- [-1, 1, Conv, [128, 3, 2, 1, 2]] # 13-P2/4
- [-1, 2, C3k2, [256, False, 0.25]]
- [-1, 1, Conv, [256, 3, 2, 1, 4]] # 15-P3/8
- [-1, 2, C3k2, [512, False, 0.25]]
- [-1, 1, Conv, [512, 3, 2]] # 17-P4/16
- [-1, 4, A2C2f, [512, True, 4]]
- [-1, 1, Conv, [1024, 3, 2]] # 19-P5/32
- [-1, 4, A2C2f, [1024, True, 1]] # 20
# YOLO12 head
head:
- [11, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 9], 1, Concat, [1]] # cat backbone P4
- [-1, 2, A2C2f, [512, False, -1]] # 23
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 7], 1, Concat, [1]] # cat backbone P3
- [-1, 2, A2C2f, [256, False, -1]] # 26
- [-1, 1, Conv, [256, 3, 2]]
- [[-1, 23], 1, Concat, [1]] # cat head P4
- [-1, 2, A2C2f, [512, False, -1]] # 29
- [-1, 1, Conv, [512, 3, 2]]
- [[-1, 11], 1, Concat, [1]] # cat head P5
- [-1, 2, C3k2, [1024, True]] # 32 (P5/32-large)
- [20, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 18], 1, Concat, [1]] # cat backbone P4
- [-1, 2, A2C2f, [512, False, -1]] # 35
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [[-1, 16], 1, Concat, [1]] # cat backbone P3
- [-1, 2, A2C2f, [256, False, -1]] # 38
- [-1, 1, Conv, [256, 3, 2]]
- [[-1, 35], 1, Concat, [1]] # cat head P4
- [-1, 2, A2C2f, [512, False, -1]] # 41
- [-1, 1, Conv, [512, 3, 2]]
- [[-1, 20], 1, Concat, [1]] # cat head P5
- [-1, 2, C3k2, [1024, True]] # 44 (P5/32-large)
- [[26, 38], 1, CGAFusion, []] # cat head P5 45
- [[29, 41], 1, CGAFusion, []] # cat head P5 46
- [[32, 44], 1, CGAFusion, []] # cat head P5 47
- [[45, 46, 47], 1, Detect, [nc]] # Detect(P3, P4, P5)
六、成功运行结果
打印网络模型可以看到不同的融合层已经加入到模型中，并可以进行训练了。
YOLOv12-mid-CGAFusion ：
YOLOv12-mid-CGAFusion summary: 752 layers, 4,071,548 parameters, 4,071,532 gradients, 9.4 GFLOPs
from n params module arguments
0 -1 1 0 ultralytics.nn.AddModules.multimodal.IN []
1 -1 1 0 ultralytics.nn.AddModules.multimodal.Multiin [1]
2 -2 1 0 ultralytics.nn.AddModules.multimodal.Multiin [2]
3 1 1 464 ultralytics.nn.modules.conv.Conv [3, 16, 3, 2]
4 -1 1 2368 ultralytics.nn.modules.conv.Conv [16, 32, 3, 2, 1, 2]
5 -1 1 6640 ultralytics.nn.modules.block.C3k2 [32, 64, 1, False, 0.25]
6 -1 1 9344 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2, 1, 4]
7 -1 1 26080 ultralytics.nn.modules.block.C3k2 [64, 128, 1, False, 0.25]
8 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
9 -1 2 180864 ultralytics.nn.AddModules.A2C2f.A2C2f [128, 128, 2, True, 4]
10 -1 1 295424 ultralytics.nn.modules.conv.Conv [128, 256, 3, 2]
11 -1 2 689408 ultralytics.nn.AddModules.A2C2f.A2C2f [256, 256, 2, True, 1]
12 2 1 464 ultralytics.nn.modules.conv.Conv [3, 16, 3, 2]
13 -1 1 2368 ultralytics.nn.modules.conv.Conv [16, 32, 3, 2, 1, 2]
14 -1 1 6640 ultralytics.nn.modules.block.C3k2 [32, 64, 1, False, 0.25]
15 -1 1 9344 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2, 1, 4]
16 -1 1 26080 ultralytics.nn.modules.block.C3k2 [64, 128, 1, False, 0.25]
17 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
18 -1 2 180864 ultralytics.nn.AddModules.A2C2f.A2C2f [128, 128, 2, True, 4]
19 -1 1 295424 ultralytics.nn.modules.conv.Conv [128, 256, 3, 2]
20 -1 2 689408 ultralytics.nn.AddModules.A2C2f.A2C2f [256, 256, 2, True, 1]
21 [7, 16] 1 33523 ultralytics.nn.AddModules.CGAFusion.CGAFusion[128]
22 [9, 18] 1 33523 ultralytics.nn.AddModules.CGAFusion.CGAFusion[128]
23 [11, 20] 1 107907 ultralytics.nn.AddModules.CGAFusion.CGAFusion[256]
24 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
25 [-1, 22] 1 0 ultralytics.nn.modules.conv.Concat [1]
26 -1 1 86912 ultralytics.nn.AddModules.A2C2f.A2C2f [384, 128, 1, False, -1]
27 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
28 [-1, 21] 1 0 ultralytics.nn.modules.conv.Concat [1]
29 -1 1 24000 ultralytics.nn.AddModules.A2C2f.A2C2f [256, 64, 1, False, -1]
30 -1 1 36992 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2]
31 [-1, 26] 1 0 ultralytics.nn.modules.conv.Concat [1]
32 -1 1 74624 ultralytics.nn.AddModules.A2C2f.A2C2f [192, 128, 1, False, -1]
33 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
34 [-1, 23] 1 0 ultralytics.nn.modules.conv.Concat [1]
35 -1 1 378880 ultralytics.nn.modules.block.C3k2 [384, 256, 1, True]
36 [29, 32, 35] 1 430867 ultralytics.nn.modules.head.Detect [1, [64, 128, 256]]
YOLOv12-mid-CGAFusion summary: 752 layers, 4,071,548 parameters, 4,071,532 gradients, 9.4 GFLOPs
YOLOv12-mid-to-late-CGAFusion ：
YOLOv12-mid-to-late-CGAFusion summary: 822 layers, 4,160,628 parameters, 4,160,612 gradients, 9.8 GFLOPs
from n params module arguments
0 -1 1 0 ultralytics.nn.AddModules.multimodal.IN []
1 -1 1 0 ultralytics.nn.AddModules.multimodal.Multiin [1]
2 -2 1 0 ultralytics.nn.AddModules.multimodal.Multiin [2]
3 1 1 464 ultralytics.nn.modules.conv.Conv [3, 16, 3, 2]
4 -1 1 2368 ultralytics.nn.modules.conv.Conv [16, 32, 3, 2, 1, 2]
5 -1 1 6640 ultralytics.nn.modules.block.C3k2 [32, 64, 1, False, 0.25]
6 -1 1 9344 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2, 1, 4]
7 -1 1 26080 ultralytics.nn.modules.block.C3k2 [64, 128, 1, False, 0.25]
8 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
9 -1 2 180864 ultralytics.nn.AddModules.A2C2f.A2C2f [128, 128, 2, True, 4]
10 -1 1 295424 ultralytics.nn.modules.conv.Conv [128, 256, 3, 2]
11 -1 2 689408 ultralytics.nn.AddModules.A2C2f.A2C2f [256, 256, 2, True, 1]
12 2 1 464 ultralytics.nn.modules.conv.Conv [3, 16, 3, 2]
13 -1 1 2368 ultralytics.nn.modules.conv.Conv [16, 32, 3, 2, 1, 2]
14 -1 1 6640 ultralytics.nn.modules.block.C3k2 [32, 64, 1, False, 0.25]
15 -1 1 9344 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2, 1, 4]
16 -1 1 26080 ultralytics.nn.modules.block.C3k2 [64, 128, 1, False, 0.25]
17 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
18 -1 2 180864 ultralytics.nn.AddModules.A2C2f.A2C2f [128, 128, 2, True, 4]
19 -1 1 295424 ultralytics.nn.modules.conv.Conv [128, 256, 3, 2]
20 -1 2 689408 ultralytics.nn.AddModules.A2C2f.A2C2f [256, 256, 2, True, 1]
21 11 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
22 [-1, 9] 1 0 ultralytics.nn.modules.conv.Concat [1]
23 -1 1 86912 ultralytics.nn.AddModules.A2C2f.A2C2f [384, 128, 1, False, -1]
24 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
25 [-1, 7] 1 0 ultralytics.nn.modules.conv.Concat [1]
26 -1 1 24000 ultralytics.nn.AddModules.A2C2f.A2C2f [256, 64, 1, False, -1]
27 20 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
28 [-1, 18] 1 0 ultralytics.nn.modules.conv.Concat [1]
29 -1 1 86912 ultralytics.nn.AddModules.A2C2f.A2C2f [384, 128, 1, False, -1]
30 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
31 [-1, 16] 1 0 ultralytics.nn.modules.conv.Concat [1]
32 -1 1 24000 ultralytics.nn.AddModules.A2C2f.A2C2f [256, 64, 1, False, -1]
33 [11, 20] 1 107907 ultralytics.nn.AddModules.CGAFusion.CGAFusion[256]
34 [23, 29] 1 33523 ultralytics.nn.AddModules.CGAFusion.CGAFusion[128]
35 [26, 32] 1 11691 ultralytics.nn.AddModules.CGAFusion.CGAFusion[64]
36 -1 1 36992 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2]
37 [-1, 34] 1 0 ultralytics.nn.modules.conv.Concat [1]
38 -1 1 74624 ultralytics.nn.AddModules.A2C2f.A2C2f [192, 128, 1, False, -1]
39 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
40 [-1, 33] 1 0 ultralytics.nn.modules.conv.Concat [1]
41 -1 1 378880 ultralytics.nn.modules.block.C3k2 [384, 256, 1, True]
42 [35, 38, 41] 1 430867 ultralytics.nn.modules.head.Detect [1, [64, 128, 256]]
YOLOv12-mid-to-late-CGAFusion summary: 822 layers, 4,160,628 parameters, 4,160,612 gradients, 9.8 GFLOPs
YOLOv12-late-CGAFusion ：
YOLOv8-late-CGAFusion summary: 463 layers, 5,103,060 parameters, 5,103,044 gradients, 12.5 GFLOPs
from n params module arguments
0 -1 1 0 ultralytics.nn.AddModules.multimodal.IN []
1 -1 1 0 ultralytics.nn.AddModules.multimodal.Multiin [1]
2 -2 1 0 ultralytics.nn.AddModules.multimodal.Multiin [2]
3 1 1 464 ultralytics.nn.modules.conv.Conv [3, 16, 3, 2]
4 -1 1 4672 ultralytics.nn.modules.conv.Conv [16, 32, 3, 2]
5 -1 1 7360 ultralytics.nn.modules.block.C2f [32, 32, 1, True]
6 -1 1 18560 ultralytics.nn.modules.conv.Conv [32, 64, 3, 2]
7 -1 2 49664 ultralytics.nn.modules.block.C2f [64, 64, 2, True]
8 -1 1 73984 ultralytics.nn.modules.conv.Conv [64, 128, 3, 2]
9 -1 2 197632 ultralytics.nn.modules.block.C2f [128, 128, 2, True]
10 -1 1 295424 ultralytics.nn.modules.conv.Conv [128, 256, 3, 2]
11 -1 1 460288 ultralytics.nn.modules.block.C2f [256, 256, 1, True]
12 -1 1 164608 ultralytics.nn.modules.block.SPPF [256, 256, 5]
13 2 1 464 ultralytics.nn.modules.conv.Conv [3, 16, 3, 2]
14 -1 1 4672 ultralytics.nn.modules.conv.Conv [16, 32, 3, 2]
15 -1 1 7360 ultralytics.nn.modules.block.C2f [32, 32, 1, True]
16 -1 1 18560 ultralytics.nn.modules.conv.Conv [32, 64, 3, 2]
17 -1 2 49664 ultralytics.nn.modules.block.C2f [64, 64, 2, True]
18 -1 1 73984 ultralytics.nn.modules.conv.Conv [64, 128, 3, 2]
19 -1 2 197632 ultralytics.nn.modules.block.C2f [128, 128, 2, True]
20 -1 1 295424 ultralytics.nn.modules.conv.Conv [128, 256, 3, 2]
21 -1 1 460288 ultralytics.nn.modules.block.C2f [256, 256, 1, True]
22 -1 1 164608 ultralytics.nn.modules.block.SPPF [256, 256, 5]
23 12 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
24 [-1, 9] 1 0 ultralytics.nn.modules.conv.Concat [1]
25 -1 1 148224 ultralytics.nn.modules.block.C2f [384, 128, 1]
26 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
27 [-1, 7] 1 0 ultralytics.nn.modules.conv.Concat [1]
28 -1 1 37248 ultralytics.nn.modules.block.C2f [192, 64, 1]
29 -1 1 36992 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2]
30 [-1, 25] 1 0 ultralytics.nn.modules.conv.Concat [1]
31 -1 1 123648 ultralytics.nn.modules.block.C2f [192, 128, 1]
32 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
33 [-1, 12] 1 0 ultralytics.nn.modules.conv.Concat [1]
34 -1 1 493056 ultralytics.nn.modules.block.C2f [384, 256, 1]
35 22 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
36 [-1, 19] 1 0 ultralytics.nn.modules.conv.Concat [1]
37 -1 1 148224 ultralytics.nn.modules.block.C2f [384, 128, 1]
38 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
39 [-1, 17] 1 0 ultralytics.nn.modules.conv.Concat [1]
40 -1 1 37248 ultralytics.nn.modules.block.C2f [192, 64, 1]
41 -1 1 36992 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2]
42 [-1, 37] 1 0 ultralytics.nn.modules.conv.Concat [1]
43 -1 1 123648 ultralytics.nn.modules.block.C2f [192, 128, 1]
44 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
45 [-1, 22] 1 0 ultralytics.nn.modules.conv.Concat [1]
46 -1 1 493056 ultralytics.nn.modules.block.C2f [384, 256, 1]
47 [28, 40] 1 11691 ultralytics.nn.AddModules.CGAFusion.CGAFusion[64]
48 [31, 43] 1 33523 ultralytics.nn.AddModules.CGAFusion.CGAFusion[128]
49 [34, 46] 1 107907 ultralytics.nn.AddModules.CGAFusion.CGAFusion[256]
50 [47, 48, 49] 1 430867 ultralytics.nn.modules.head.Detect [1, [64, 128, 256]]
YOLOv8-late-CGAFusion summary: 463 layers, 5,103,060 parameters, 5,103,044 gradients, 12.5 GFLOPs