RT-DETR改进策略【卷积层】| NeurIPS-2022 ParNet 即插即用模块 二次创新ResNetLayer
一、本文介绍
本文记录的是
利用
ParNet
中的基础模块优化
RT-DETR
的目标检测网络模型
。
ParNet block
是一个即插即用模块，
能够在不增加深度的情况下增加感受野，更好地处理图像中的不同尺度特征，有助于网络对输入数据更全面地理解和学习，从而提升网络的特征提取能力和分类性能。
二、ParNet 介绍
NON-DEEP NETWORKS
ParNet block
模块是ParNet网络架构中的重要组成部分，其设计出发点、原理、结构和优势如下：
2,1 设计出发点
- 探索非深度网络的高性能 ：为了研究是否可以通过非深度（约10层）的神经网络在竞争激烈的基准测试中实现高性能，需要一种适合的网络模块结构。
- 解决训练难题与提升性能 ：VGG-style网络训练通常比ResNet-style网络更困难，但通过“结构重参数化”技术可以使训练变得容易些，同时还需要解决非深度网络中如 receptive field有限以及可能缺乏足够非线性等问题，以提升性能。
2.2 原理
- 结构重参数化 ：在训练过程中使用多个分支的 3 × 3 3×3 3 × 3 卷积块，训练完成后将这些分支融合成一个 3 × 3 3×3 3 × 3 卷积，从而减少推理过程中的延迟。
- 增加非线性与提升感受野 ：采用Skip - Squeeze - Excitation (SSE) 层来增加感受野且不影响深度，同时用SiLU激活函数替换ReLU激活函数以增加网络的非线性。
2.3 结构
- 基础结构 ：从Rep - VGG块借鉴初始设计并进行修改。
- SSE模块 ：基于Squeeze - and - Excitation (SE) 设计构建Skip - Squeeze - Excitation (SSE) 层，该层与跳跃连接一起使用，并包含一个单一的全连接层。
- 激活函数 ：采用SiLU激活函数。
论文： https://arxiv.org/pdf/2110.07641
源码： https://github.com/imankgoyal/NonDeepNetworks
三、ParNetblock的实现代码
ParNetblock模块
的实现代码如下：
import torch
from torch import nn
import torch.nn.functional as F
class ParNetblock(nn.Module):
def __init__(self, channel, out_channel):
super().__init__()
self.sse = nn.Sequential(
nn.AdaptiveAvgPool2d(1),
nn.Conv2d(channel, out_channel, kernel_size=1),
nn.Sigmoid())
self.conv1x1 = nn.Sequential(
nn.Conv2d(channel, out_channel, kernel_size=1),
nn.BatchNorm2d(out_channel))
self.conv3x3 = nn.Sequential(
nn.Conv2d(channel, out_channel, kernel_size=3, padding=1),
nn.BatchNorm2d(out_channel))
self.silu = nn.SiLU()
def forward(self, x):
b, c, _, _ = x.size()
x1 = self.conv1x1(x)
x2 = self.conv3x3(x)
x3 = self.sse(x) * x
y = self.silu(x1 + x2 + x3)
return y
def autopad(k, p=None, d=1):
"""
Pads kernel to 'same' output shape, adjusting for optional dilation; returns padding size.
`k`: kernel, `p`: padding, `d`: dilation.
"""
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):
"""Initializes a standard convolution layer with optional batch normalization and 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):
"""Applies a convolution followed by batch normalization and an activation function to the input tensor `x`."""
return self.act(self.bn(self.conv(x)))
def forward_fuse(self, x):
"""Applies a fused convolution and activation function to the input tensor `x`."""
return self.act(self.conv(x))
class ResNetBlock(nn.Module):
"""ResNet block with standard convolution layers."""
def __init__(self, c1, c2, s=1, e=4):
"""Initialize convolution with given parameters."""
super().__init__()
c3 = e * c2
self.cv1 = Conv(c1, c2, k=1, s=1, act=True)
self.cv2 = Conv(c2, c2, k=3, s=s, p=1, act=True)
self.cv3 = Conv(c2, c3, k=1, act=False)
self.cv4 = ParNetblock(c2, c2)
self.shortcut = nn.Sequential(Conv(c1, c3, k=1, s=s, act=False)) if s != 1 or c1 != c3 else nn.Identity()
def forward(self, x):
"""Forward pass through the ResNet block."""
return F.relu(self.cv3(self.cv4(self.cv2(self.cv1(x)))) + self.shortcut(x))
class ResNetLayer_ParNet(nn.Module):
"""ResNet layer with multiple ResNet blocks."""
def __init__(self, c1, c2, s=1, is_first=False, n=1, e=4):
"""Initializes the ResNetLayer given arguments."""
super().__init__()
self.is_first = is_first
if self.is_first:
self.layer = nn.Sequential(
Conv(c1, c2, k=7, s=2, p=3, act=True), nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
)
else:
blocks = [ResNetBlock(c1, c2, s, e=e)]
blocks.extend([ResNetBlock(e * c2, c2, 1, e=e) for _ in range(n - 1)])
self.layer = nn.Sequential(*blocks)
def forward(self, x):
"""Forward pass through the ResNet layer."""
return self.layer(x)
四、创新模块
4.1 改进点1⭐
模块改进方法
：直接加入
ParNetblock模块
（
第五节讲解添加步骤
）。
ParNetblock模块
添加后如下：
4.2 改进点2⭐
模块改进方法
：基于
ParNetblock模块
的
ResNetLayer
（
第五节讲解添加步骤
）。
第二种改进方法是对
RT-DETR
中的
ResNetLayer模块
进行改进，并将
ParNetblock
在加入到
ResNetLayer
模块中。
改进代码如下：
首先添加如下代码改进
ResNetBlock
模块，并将
ResNetLayer
重命名为
ResNetLayer_ParNet
class ResNetBlock(nn.Module):
"""ResNet block with standard convolution layers."""
def __init__(self, c1, c2, s=1, e=4):
"""Initialize convolution with given parameters."""
super().__init__()
c3 = e * c2
self.cv1 = Conv(c1, c2, k=1, s=1, act=True)
self.cv2 = Conv(c2, c2, k=3, s=s, p=1, act=True)
self.cv3 = Conv(c2, c3, k=1, act=False)
self.cv4 = ParNetblock(c2, c2)
self.shortcut = nn.Sequential(Conv(c1, c3, k=1, s=s, act=False)) if s != 1 or c1 != c3 else nn.Identity()
def forward(self, x):
"""Forward pass through the ResNet block."""
return F.relu(self.cv3(self.cv4(self.cv2(self.cv1(x)))) + self.shortcut(x))
class ResNetLayer_ParNet(nn.Module):
"""ResNet layer with multiple ResNet blocks."""
def __init__(self, c1, c2, s=1, is_first=False, n=1, e=4):
"""Initializes the ResNetLayer given arguments."""
super().__init__()
self.is_first = is_first
if self.is_first:
self.layer = nn.Sequential(
Conv(c1, c2, k=7, s=2, p=3, act=True), nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
)
else:
blocks = [ResNetBlock(c1, c2, s, e=e)]
blocks.extend([ResNetBlock(e * c2, c2, 1, e=e) for _ in range(n - 1)])
self.layer = nn.Sequential(*blocks)
def forward(self, x):
"""Forward pass through the ResNet layer."""
return self.layer(x)
注意❗：在
第五小节
中需要声明的模块名称为：
ResNetLayer_ParNet
。
五、添加步骤
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
ParNet.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 .ParNet 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
ParNetblock
和
ResNetLayer_ParNet
模块
六、yaml模型文件
6.1 模型改进版本1
此处以
ultralytics/cfg/models/rt-detr/rtdetr-l.yaml
为例，在同目录下创建一个用于自己数据集训练的模型文件
rtdetr-l-ParNetblock.yaml
。
将
rtdetr-l.yaml
中的内容复制到
rtdetr-l-ParNetblock.yaml
文件下，修改
nc
数量等于自己数据中目标的数量。
📌 模型的修改方法是将
骨干网络
中添加
ParNetblock模块
。
# Ultralytics YOLO 🚀, AGPL-3.0 license
# RT-DETR-l object detection model with P3-P5 outputs. For details see https://docs.ultralytics.com/models/rtdetr
# Parameters
nc: 1 # number of classes
scales: # model compound scaling constants, i.e. 'model=yolov8n-cls.yaml' will call yolov8-cls.yaml with scale 'n'
# [depth, width, max_channels]
l: [1.00, 1.00, 1024]
backbone:
# [from, repeats, module, args]
- [-1, 1, HGStem, [32, 48]] # 0-P2/4
- [-1, 6, HGBlock, [48, 128, 3]] # stage 1
- [-1, 1, DWConv, [128, 3, 2, 1, False]] # 2-P3/8
- [-1, 6, HGBlock, [96, 512, 3]] # stage 2
- [-1, 1, DWConv, [512, 3, 2, 1, False]] # 4-P4/16
- [-1, 6, ParNetblock, [512]] # cm, c2, k, light, shortcut
- [-1, 6, ParNetblock, [512]]
- [-1, 6, ParNetblock, [512]] # stage 3
- [-1, 1, DWConv, [1024, 3, 2, 1, False]] # 8-P5/32
- [-1, 6, HGBlock, [384, 2048, 5, True, False]] # stage 4
head:
- [-1, 1, Conv, [256, 1, 1, None, 1, 1, False]] # 10 input_proj.2
- [-1, 1, AIFI, [1024, 8]]
- [-1, 1, Conv, [256, 1, 1]] # 12, Y5, lateral_convs.0
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [7, 1, Conv, [256, 1, 1, None, 1, 1, False]] # 14 input_proj.1
- [[-2, -1], 1, Concat, [1]]
- [-1, 3, RepC3, [256]] # 16, fpn_blocks.0
- [-1, 1, Conv, [256, 1, 1]] # 17, Y4, lateral_convs.1
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [3, 1, Conv, [256, 1, 1, None, 1, 1, False]] # 19 input_proj.0
- [[-2, -1], 1, Concat, [1]] # cat backbone P4
- [-1, 3, RepC3, [256]] # X3 (21), fpn_blocks.1
- [-1, 1, Conv, [256, 3, 2]] # 22, downsample_convs.0
- [[-1, 17], 1, Concat, [1]] # cat Y4
- [-1, 3, RepC3, [256]] # F4 (24), pan_blocks.0
- [-1, 1, Conv, [256, 3, 2]] # 25, downsample_convs.1
- [[-1, 12], 1, Concat, [1]] # cat Y5
- [-1, 3, RepC3, [256]] # F5 (27), pan_blocks.1
- [[21, 24, 27], 1, RTDETRDecoder, [nc]] # Detect(P3, P4, P5)
6.2 模型改进版本2⭐
此处以
ultralytics/cfg/models/rt-detr/rtdetr-resnet50.yaml
为例，在同目录下创建一个用于自己数据集训练的模型文件
rtdetr-ResNetLayer_ParNet.yaml
。
将
rtdetr-resnet50.yaml
中的内容复制到
rtdetr-ResNetLayer_ParNet.yaml
文件下，修改
nc
数量等于自己数据中目标的数量。
📌 模型的修改方法是将
骨干网络
中的所有
ResNetLayer模块
替换成
ResNetLayer_ParNet模块
。
# Ultralytics YOLO 🚀, AGPL-3.0 license
# RT-DETR-ResNet50 object detection model with P3-P5 outputs.
# Parameters
nc: 1 # number of classes
scales: # model compound scaling constants, i.e. 'model=yolov8n-cls.yaml' will call yolov8-cls.yaml with scale 'n'
# [depth, width, max_channels]
l: [1.00, 1.00, 1024]
backbone:
# [from, repeats, module, args]
- [-1, 1, ResNetLayer_ParNet, [3, 64, 1, True, 1]] # 0
- [-1, 1, ResNetLayer_ParNet, [64, 64, 1, False, 3]] # 1
- [-1, 1, ResNetLayer_ParNet, [256, 128, 2, False, 4]] # 2
- [-1, 1, ResNetLayer_ParNet, [512, 256, 2, False, 6]] # 3
- [-1, 1, ResNetLayer_ParNet, [1024, 512, 2, False, 3]] # 4
head:
- [-1, 1, Conv, [256, 1, 1, None, 1, 1, False]] # 5
- [-1, 1, AIFI, [1024, 8]]
- [-1, 1, Conv, [256, 1, 1]] # 7
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [3, 1, Conv, [256, 1, 1, None, 1, 1, False]] # 9
- [[-2, -1], 1, Concat, [1]]
- [-1, 3, RepC3, [256]] # 11
- [-1, 1, Conv, [256, 1, 1]] # 12
- [-1, 1, nn.Upsample, [None, 2, "nearest"]]
- [2, 1, Conv, [256, 1, 1, None, 1, 1, False]] # 14
- [[-2, -1], 1, Concat, [1]] # cat backbone P4
- [-1, 3, RepC3, [256]] # X3 (16), fpn_blocks.1
- [-1, 1, Conv, [256, 3, 2]] # 17, downsample_convs.0
- [[-1, 12], 1, Concat, [1]] # cat Y4
- [-1, 3, RepC3, [256]] # F4 (19), pan_blocks.0
- [-1, 1, Conv, [256, 3, 2]] # 20, downsample_convs.1
- [[-1, 7], 1, Concat, [1]] # cat Y5
- [-1, 3, RepC3, [256]] # F5 (22), pan_blocks.1
- [[16, 19, 22], 1, RTDETRDecoder, [nc]] # Detect(P3, P4, P5)
七、成功运行结果
打印网络模型可以看到
ParNetblock
和
C3k2_ParNet
已经加入到模型中，并可以进行训练了。
rtdetr-l-ParNetblock ：
rtdetr-l-ParNetblock summary: 732 layers, 78,839,107 parameters, 78,839,107 gradients, 240.2 GFLOPs
from n params module arguments
0 -1 1 25248 ultralytics.nn.modules.block.HGStem [3, 32, 48]
1 -1 6 155072 ultralytics.nn.modules.block.HGBlock [48, 48, 128, 3, 6]
2 -1 1 1408 ultralytics.nn.modules.conv.DWConv [128, 128, 3, 2, 1, False]
3 -1 6 839296 ultralytics.nn.modules.block.HGBlock [128, 96, 512, 3, 6]
4 -1 1 5632 ultralytics.nn.modules.conv.DWConv [512, 512, 3, 2, 1, False]
5 -1 6 17323008 ultralytics.nn.AddModules.ParNet.ParNetblock [512, 512]
6 -1 6 17323008 ultralytics.nn.AddModules.ParNet.ParNetblock [512, 512]
7 -1 6 17323008 ultralytics.nn.AddModules.ParNet.ParNetblock [512, 512]
8 -1 1 11264 ultralytics.nn.modules.conv.DWConv [512, 1024, 3, 2, 1, False]
9 -1 6 6708480 ultralytics.nn.modules.block.HGBlock [1024, 384, 2048, 5, 6, True, False]
10 -1 1 524800 ultralytics.nn.modules.conv.Conv [2048, 256, 1, 1, None, 1, 1, False]
11 -1 1 789760 ultralytics.nn.modules.transformer.AIFI [256, 1024, 8]
12 -1 1 66048 ultralytics.nn.modules.conv.Conv [256, 256, 1, 1]
13 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
14 7 1 131584 ultralytics.nn.modules.conv.Conv [512, 256, 1, 1, None, 1, 1, False]
15 [-2, -1] 1 0 ultralytics.nn.modules.conv.Concat [1]
16 -1 3 2232320 ultralytics.nn.modules.block.RepC3 [512, 256, 3]
17 -1 1 66048 ultralytics.nn.modules.conv.Conv [256, 256, 1, 1]
18 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
19 3 1 131584 ultralytics.nn.modules.conv.Conv [512, 256, 1, 1, None, 1, 1, False]
20 [-2, -1] 1 0 ultralytics.nn.modules.conv.Concat [1]
21 -1 3 2232320 ultralytics.nn.modules.block.RepC3 [512, 256, 3]
22 -1 1 590336 ultralytics.nn.modules.conv.Conv [256, 256, 3, 2]
23 [-1, 17] 1 0 ultralytics.nn.modules.conv.Concat [1]
24 -1 3 2232320 ultralytics.nn.modules.block.RepC3 [512, 256, 3]
25 -1 1 590336 ultralytics.nn.modules.conv.Conv [256, 256, 3, 2]
26 [-1, 12] 1 0 ultralytics.nn.modules.conv.Concat [1]
27 -1 3 2232320 ultralytics.nn.modules.block.RepC3 [512, 256, 3]
28 [21, 24, 27] 1 7303907 ultralytics.nn.modules.head.RTDETRDecoder [1, [256, 256, 256]]
rtdetr-l-ParNetblock summary: 732 layers, 78,839,107 parameters, 78,839,107 gradients, 240.2 GFLOPs
rtdetr-ResNetLayer_ParNet ：
rtdetr-ResNetLayer_ParNet summary: 785 layers, 56,621,411 parameters, 56,621,411 gradients, 164.2 GFLOPs
from n params module arguments
0 -1 1 9536 ultralytics.nn.AddModules.ParNet.ResNetLayer_ParNet[3, 64, 1, True, 1]
1 -1 1 352320 ultralytics.nn.AddModules.ParNet.ResNetLayer_ParNet[64, 64, 1, False, 3]
2 -1 1 1944064 ultralytics.nn.AddModules.ParNet.ResNetLayer_ParNet[256, 128, 2, False, 4]
3 -1 1 11434496 ultralytics.nn.AddModules.ParNet.ResNetLayer_ParNet[512, 256, 2, False, 6]
4 -1 1 23626240 ultralytics.nn.AddModules.ParNet.ResNetLayer_ParNet[1024, 512, 2, False, 3]
5 -1 1 524800 ultralytics.nn.modules.conv.Conv [2048, 256, 1, 1, None, 1, 1, False]
6 -1 1 789760 ultralytics.nn.modules.transformer.AIFI [256, 1024, 8]
7 -1 1 66048 ultralytics.nn.modules.conv.Conv [256, 256, 1, 1]
8 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
9 3 1 262656 ultralytics.nn.modules.conv.Conv [1024, 256, 1, 1, None, 1, 1, False]
10 [-2, -1] 1 0 ultralytics.nn.modules.conv.Concat [1]
11 -1 3 2232320 ultralytics.nn.modules.block.RepC3 [512, 256, 3]
12 -1 1 66048 ultralytics.nn.modules.conv.Conv [256, 256, 1, 1]
13 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, 'nearest']
14 2 1 131584 ultralytics.nn.modules.conv.Conv [512, 256, 1, 1, None, 1, 1, False]
15 [-2, -1] 1 0 ultralytics.nn.modules.conv.Concat [1]
16 -1 3 2232320 ultralytics.nn.modules.block.RepC3 [512, 256, 3]
17 -1 1 590336 ultralytics.nn.modules.conv.Conv [256, 256, 3, 2]
18 [-1, 12] 1 0 ultralytics.nn.modules.conv.Concat [1]
19 -1 3 2232320 ultralytics.nn.modules.block.RepC3 [512, 256, 3]
20 -1 1 590336 ultralytics.nn.modules.conv.Conv [256, 256, 3, 2]
21 [-1, 7] 1 0 ultralytics.nn.modules.conv.Concat [1]
22 -1 3 2232320 ultralytics.nn.modules.block.RepC3 [512, 256, 3]
23 [16, 19, 22] 1 7303907 ultralytics.nn.modules.head.RTDETRDecoder [1, [256, 256, 256]]
rtdetr-ResNetLayer_ParNet summary: 785 layers, 56,621,411 parameters, 56,621,411 gradients, 164.2 GFLOPs