本篇文章将介绍一个新的改进机制——WTConv（小波卷积），并阐述如何将其应用于YOLOv11中，显著提升模型性能。YOLOv11模型相比较于前几个模型在检测精度和速度上有显著提升，但其仍然受卷积核感受野大小的限制。因此，我们引入了小波卷积模块，旨在扩大卷积的感受野并有效捕捉图像中的低频信息。其对多尺度问题和小目标问题上有很好的效果。
首先，我们将解析WTConv的工作原理，它通过小波变换将输入图像分解为不同的频率成分，并在每个频率层上进行小尺寸卷积，最后通过逆小波变换将结果重新组合，从而实现对图像的多尺度分析。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. Wavelet Convolutions(WTConv)结构介绍
1. 首先，WTConv利用二维Haar小波变换对输入图像进行多级分解。Haar小波变换使用四个滤波器，将图像分解为四个子带：低频分量 (LL)：捕捉图像的低频信息，如整体形状或轮廓。水平高频分量 (LH)：捕捉图像中的水平边缘信息。垂直高频分量 (HL)：捕捉图像中的垂直边缘信息。对角线高频分量 (HH)：捕捉图像的对角线细节。在每一级的小波变换中，图像被下采样（空间分辨率减半），但频率信息得到了更细的分解。递归地执行小波变换（称为多级分解）可以得到不同尺度下的频率分量。
2. 在WTConv中，卷积并非直接在原始输入图像上进行，而是分别在每个频率子带上应用小尺寸的深度卷积核。通常使用3x3或5x5的小卷积核，对分解后的各个子带进行卷积操作。由于小波变换降低了每个子带的空间分辨率，小尺寸的卷积核可以覆盖更大的原始图像区域，即感受野增大。
低频子带 (LL) 主要包含了图像的大尺度信息，因此在其上应用卷积有助于捕捉全局的特征。
高频子带 (LH, HL, HH) 则包含局部边缘和细节信息，卷积操作可以捕捉这些细节。
3. 在完成卷积后，使用逆小波变换（Inverse Wavelet Transform, IWT）将各个子带的卷积结果重新合成为一个完整的输出。这一过程类似于将不同频率层次的特征融合在一起。值得注意的是，IWT操作是线性的，因此可以无损地将卷积结果重构到原始空间。
2. YOLOv11与WTConv的结合
1. 改进C3k2：本文使用WTConv卷积改进C3k2，构建C3k2_WT模块，然后使用C3k2_WT替换原有的C3k2，这样就可以利用WTConv扩大模型的感受野。
2. 在head中使用DSConvWithWT卷积：本文使用WTConv卷积构建DSConvWithWT深度可分离卷积，然后将它替换head中的深度可分离卷积，扩大head层的感受野。
3. Wavelet Convolutions(WTConv)代码部分
import pywt
import pywt.data
import torch
from torch import nn
from functools import partial
import torch.nn.functional as Ffrom .conv import Conv
from .block import C2f, C3, Bottleneckdef create_wavelet_filter(wave, in_size, out_size, type=torch.float):w = pywt.Wavelet(wave)dec_hi = torch.tensor(w.dec_hi[::-1], dtype=type)dec_lo = torch.tensor(w.dec_lo[::-1], dtype=type)dec_filters = torch.stack([dec_lo.unsqueeze(0) * dec_lo.unsqueeze(1),dec_lo.unsqueeze(0) * dec_hi.unsqueeze(1),dec_hi.unsqueeze(0) * dec_lo.unsqueeze(1),dec_hi.unsqueeze(0) * dec_hi.unsqueeze(1)], dim=0)dec_filters = dec_filters[:, None].repeat(in_size, 1, 1, 1)rec_hi = torch.tensor(w.rec_hi[::-1], dtype=type).flip(dims=[0])rec_lo = torch.tensor(w.rec_lo[::-1], dtype=type).flip(dims=[0])rec_filters = torch.stack([rec_lo.unsqueeze(0) * rec_lo.unsqueeze(1),rec_lo.unsqueeze(0) * rec_hi.unsqueeze(1),rec_hi.unsqueeze(0) * rec_lo.unsqueeze(1),rec_hi.unsqueeze(0) * rec_hi.unsqueeze(1)], dim=0)rec_filters = rec_filters[:, None].repeat(out_size, 1, 1, 1)return dec_filters, rec_filtersdef wavelet_transform(x, filters):b, c, h, w = x.shapepad = (filters.shape[2] // 2 - 1, filters.shape[3] // 2 - 1)x = F.conv2d(x, filters, stride=2, groups=c, padding=pad)x = x.reshape(b, c, 4, h // 2, w // 2)return xdef inverse_wavelet_transform(x, filters):b, c, _, h_half, w_half = x.shapepad = (filters.shape[2] // 2 - 1, filters.shape[3] // 2 - 1)x = x.reshape(b, c * 4, h_half, w_half)x = F.conv_transpose2d(x, filters, stride=2, groups=c, padding=pad)return x# Wavelet Transform Conv(WTConv2d)
class WTConv2d(nn.Module):def __init__(self, in_channels, out_channels, kernel_size=5, stride=1, bias=True, wt_levels=1, wt_type='db1'):super(WTConv2d, self).__init__()assert in_channels == out_channelsself.in_channels = in_channelsself.wt_levels = wt_levelsself.stride = strideself.dilation = 1self.wt_filter, self.iwt_filter = create_wavelet_filter(wt_type, in_channels, in_channels, torch.float)self.wt_filter = nn.Parameter(self.wt_filter, requires_grad=False)self.iwt_filter = nn.Parameter(self.iwt_filter, requires_grad=False)self.wt_function = partial(wavelet_transform, filters=self.wt_filter)self.iwt_function = partial(inverse_wavelet_transform, filters=self.iwt_filter)self.base_conv = nn.Conv2d(in_channels, in_channels, kernel_size, padding='same', stride=1, dilation=1,groups=in_channels, bias=bias)self.base_scale = _ScaleModule([1, in_channels, 1, 1])self.wavelet_convs = nn.ModuleList([nn.Conv2d(in_channels * 4, in_channels * 4, kernel_size, padding='same', stride=1, dilation=1,groups=in_channels * 4, bias=False) for _ in range(self.wt_levels)])self.wavelet_scale = nn.ModuleList([_ScaleModule([1, in_channels * 4, 1, 1], init_scale=0.1) for _ in range(self.wt_levels)])if self.stride > 1:self.stride_filter = nn.Parameter(torch.ones(in_channels, 1, 1, 1), requires_grad=False)self.do_stride = lambda x_in: F.conv2d(x_in, self.stride_filter, bias=None, stride=self.stride,groups=in_channels)else:self.do_stride = Nonedef forward(self, x):x_ll_in_levels = []x_h_in_levels = []shapes_in_levels = []curr_x_ll = xfor i in range(self.wt_levels):curr_shape = curr_x_ll.shapeshapes_in_levels.append(curr_shape)if (curr_shape[2] % 2 > 0) or (curr_shape[3] % 2 > 0):curr_pads = (0, curr_shape[3] % 2, 0, curr_shape[2] % 2)curr_x_ll = F.pad(curr_x_ll, curr_pads)curr_x = self.wt_function(curr_x_ll)curr_x_ll = curr_x[:, :, 0, :, :]shape_x = curr_x.shapecurr_x_tag = curr_x.reshape(shape_x[0], shape_x[1] * 4, shape_x[3], shape_x[4])curr_x_tag = self.wavelet_scale[i](self.wavelet_convs[i](curr_x_tag))curr_x_tag = curr_x_tag.reshape(shape_x)x_ll_in_levels.append(curr_x_tag[:, :, 0, :, :])x_h_in_levels.append(curr_x_tag[:, :, 1:4, :, :])next_x_ll = 0for i in range(self.wt_levels - 1, -1, -1):curr_x_ll = x_ll_in_levels.pop()curr_x_h = x_h_in_levels.pop()curr_shape = shapes_in_levels.pop()curr_x_ll = curr_x_ll + next_x_llcurr_x = torch.cat([curr_x_ll.unsqueeze(2), curr_x_h], dim=2)next_x_ll = self.iwt_function(curr_x)next_x_ll = next_x_ll[:, :, :curr_shape[2], :curr_shape[3]]x_tag = next_x_llassert len(x_ll_in_levels) == 0x = self.base_scale(self.base_conv(x))x = x + x_tagif self.do_stride is not None:x = self.do_stride(x)return xclass _ScaleModule(nn.Module):def __init__(self, dims, init_scale=1.0, init_bias=0):super(_ScaleModule, self).__init__()self.dims = dimsself.weight = nn.Parameter(torch.ones(*dims) * init_scale)self.bias = Nonedef forward(self, x):return torch.mul(self.weight, x)class DSConvWithWT(nn.Module):def __init__(self, in_channels, out_channels, kernel_size=3):super(DSConvWithWT, self).__init__()# 深度卷积：使用 WTConv2d 替换 3x3 卷积self.depthwise = WTConv2d(in_channels, in_channels, kernel_size=kernel_size)# 逐点卷积：使用 1x1 卷积self.pointwise = nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, padding=0, bias=False)def forward(self, x):x = self.depthwise(x)x = self.pointwise(x)return xclass Bottleneck_WT(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 channelsself.cv1 = Conv(c1, c_, k[0], 1)self.cv2 = WTConv2d(c_, c2)self.add = shortcut and c1 == c2def 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_WT(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_WT(c_, c_, shortcut, g, k=(k, k), e=1.0) for _ in range(n)))# 在c3k=True时，使用Bottleneck_WT特征融合，为false的时候我们使用普通的Bottleneck提取特征
class C3k2_WT(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_WT(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 = DSConvWithWT(256, 128)#创建一个输入张量batch_size = 8input_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WTConv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WTConv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
将DWConv改成DSConvWithWT
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 GFLOPss: [0.50, 0.50, 1024] # summary: 319 layers, 9458752 parameters, 9458736 gradients, 21.7 GFLOPsm: [0.50, 1.00, 512] # summary: 409 layers, 20114688 parameters, 20114672 gradients, 68.5 GFLOPsl: [1.00, 1.00, 512] # summary: 631 layers, 25372160 parameters, 25372144 gradients, 87.6 GFLOPsx: [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_WT, [256, False, 0.25]]- [-1, 1, Conv, [256, 3, 2]] # 3-P3/8- [-1, 2, C3k2_WT, [512, False, 0.25]]- [-1, 1, Conv, [512, 3, 2]] # 5-P4/16- [-1, 2, C3k2_WT, [512, True]]- [-1, 1, Conv, [1024, 3, 2]] # 7-P5/32- [-1, 2, C3k2_WT, [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, [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, 13], 1, Concat, [1]] # cat head P4- [-1, 2, C3k2, [512, False]] # 19 (P4/16-medium)- [-1, 1, Conv, [512, 3, 2]]- [[-1, 10], 1, Concat, [1]] # cat head P5- [-1, 2, C3k2, [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
第二个不需要修改配置文件
第五：运行成功
from ultralytics.models import NAS, RTDETR, SAM, YOLO, FastSAM, YOLOWorldif __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_WTConv.yaml")\.load(r'D:\bilibili\model\YOLO11\ultralytics-main\yolo11n.pt') # build from YAML and transfer weightsresults = model.train(data=r'D:\bilibili\model\ultralytics-main\ultralytics\cfg\datasets\VOC_my.yaml',epochs=100, imgsz=640, batch=8) 
