YOLO11 改进 – 检测头 _ Detect-Dyhead动态检测头：轻量级设计统一多头注意力，提升复杂场景感知效率

前言

本文介绍了Detect-Dyhead检测头，并将其集成进YOLOv11。Detect-Dyhead通过将注意力功能分解为尺度意识、空间意识和任务意识三种机制，在不大幅增加计算开销的情况下，提高了模型的表现能力。我们将Detect-Dyhead的代码集成到YOLOv11中，并在tasks文件中进行注册。实验证明，YOLOv11-Detect-Dyhead在目标检测任务中取得了良好的效果。

文章目录： YOLOv11改进大全：卷积层、轻量化、注意力机制、损失函数、Backbone、SPPF、Neck、检测头全方位优化汇总

专栏链接: YOLOv11改进专栏

文章目录

[TOC]

原理介绍

参考文章：

https://www.nature.com/articles/s41598-024-67953-3

https://www.mdpi.com/2079-9292/13/1/43

用 Detect-Dyhead（Detect-Dynamic 头）替换原来的检测头，在不大幅增加计算开销的情况下，大大提高了模型的表现能力。鉴于张量的高维度，在所有维度上直接学习注意力函数被证明在计算上是昂贵的。Detect-Dyhead 通过将注意力功能分解为三种不同的机制来巧妙地应对这一挑战，每种机制都专注于一个特定的视角。这三个视角包括尺度意识、空间意识和任务意识，每种注意力机制都是独立应用的。下图 阐明了 Detect-Dyhead 网络的结果架构。

核心代码

class DyConv(nn.Module):
    def __init__(self, in_channels=256, out_channels=256, conv_func=Conv3x3Norm):
        super(DyConv, self).__init__()

        self.DyConv = nn.ModuleList()
        self.DyConv.append(conv_func(in_channels, out_channels, 1))
        self.DyConv.append(conv_func(in_channels, out_channels, 1))
        self.DyConv.append(conv_func(in_channels, out_channels, 2))

        self.AttnConv = nn.Sequential(
            nn.AdaptiveAvgPool2d(1),
            nn.Conv2d(in_channels, 1, kernel_size=1),
            nn.ReLU(inplace=True))

        self.h_sigmoid = h_sigmoid()
        self.relu = DYReLU(in_channels, out_channels)
        self.offset = nn.Conv2d(in_channels, 27, kernel_size=3, stride=1, padding=1)
        self.init_weights()

    def init_weights(self):
        for m in self.DyConv.modules():
            if isinstance(m, nn.Conv2d):
                nn.init.normal_(m.weight.data, 0, 0.01)
                if m.bias is not None:
                    m.bias.data.zero_()
        for m in self.AttnConv.modules():
            if isinstance(m, nn.Conv2d):
                nn.init.normal_(m.weight.data, 0, 0.01)
                if m.bias is not None:
                    m.bias.data.zero_()

    def forward(self, x):
        next_x = {}
        feature_names = list(x.keys())
        for level, name in enumerate(feature_names):

            feature = x[name]

            offset_mask = self.offset(feature)
            offset = offset_mask[:, :18, :, :]
            mask = offset_mask[:, 18:, :, :].sigmoid()
            conv_args = dict(offset=offset, mask=mask)

            temp_fea = [self.DyConv[1](feature, **conv_args)]
            if level > 0:
                temp_fea.append(self.DyConv[2](x[feature_names[level - 1]], **conv_args))
            if level < len(x) - 1:
                input = x[feature_names[level + 1]]
                temp_fea.append(F.interpolate(self.DyConv[0](input, **conv_args),
                                              size=[feature.size(2), feature.size(3)]))
            attn_fea = []
            res_fea = []
            for fea in temp_fea:
                res_fea.append(fea)
                attn_fea.append(self.AttnConv(fea))

            res_fea = torch.stack(res_fea)
            spa_pyr_attn = self.h_sigmoid(torch.stack(attn_fea))
            mean_fea = torch.mean(res_fea * spa_pyr_attn, dim=0, keepdim=False)
            next_x[name] = self.relu(mean_fea)

        return next_x

class Detect_DyHead(nn.Module):
    """YOLOv8 Detect head for detection models."""
    dynamic = False  # force grid reconstruction
    export = False  # export mode
    shape = None
    anchors = torch.empty(0)  # init
    strides = torch.empty(0)  # init

    def __init__(self, nc=80, ch=()):
        """Initializes the YOLOv8 detection layer with specified number of classes and channels."""
        super().__init__()
        self.nc = nc  # number of classes
        self.nl = len(ch)  # number of detection layers
        self.reg_max = 16  # DFL channels (ch[0] // 16 to scale 4/8/12/16/20 for n/s/m/l/x)
        self.no = nc + self.reg_max * 4  # number of outputs per anchor
        self.stride = torch.zeros(self.nl)  # strides computed during build
        c2, c3 = max((16, ch[0] // 4, self.reg_max * 4)), max(ch[0], min(self.nc, 100))  # channels
        self.cv2 = nn.ModuleList(
            nn.Sequential(Conv(x, c2, 3), Conv(c2, c2, 3), nn.Conv2d(c2, 4 * self.reg_max, 1)) for x in ch)
        self.cv3 = nn.ModuleList(nn.Sequential(Conv(x, c3, 3), Conv(c3, c3, 3), nn.Conv2d(c3, self.nc, 1)) for x in ch)
        self.dfl = DFL(self.reg_max) if self.reg_max > 1 else nn.Identity()
        dyhead_tower = []
        for i in range(self.nl):
            channel = ch[i]
            dyhead_tower.append(
                DyConv(
                    channel,
                    channel,
                    conv_func=Conv3x3Norm,
                )
            )
        self.add_module('dyhead_tower', nn.Sequential(*dyhead_tower))

    def forward(self, x):
        tensor_dict = {i: tensor for i, tensor in enumerate(x)}
        x = self.dyhead_tower(tensor_dict)
        x = list(x.values())
        """Concatenates and returns predicted bounding boxes and class probabilities."""
        shape = x[0].shape  # BCHW
        for i in range(self.nl):
            x[i] = torch.cat((self.cv2[i](x[i]), self.cv3[i](x[i])), 1)
        if self.training:
            return x
        elif self.dynamic or self.shape != shape:
            self.anchors, self.strides = (x.transpose(0, 1) for x in make_anchors(x, self.stride, 0.5))
            self.shape = shape

        x_cat = torch.cat([xi.view(shape[0], self.no, -1) for xi in x], 2)
        if self.export and self.format in ('saved_model', 'pb', 'tflite', 'edgetpu', 'tfjs'):  # avoid TF FlexSplitV ops
            box = x_cat[:, :self.reg_max * 4]
            cls = x_cat[:, self.reg_max * 4:]
        else:
            box, cls = x_cat.split((self.reg_max * 4, self.nc), 1)
        dbox = dist2bbox(self.dfl(box), self.anchors.unsqueeze(0), xywh=True, dim=1) * self.strides

        if self.export and self.format in ('tflite', 'edgetpu'):
            # Normalize xywh with image size to mitigate quantization error of TFLite integer models as done in YOLOv5:
            # https://github.com/ultralytics/yolov5/blob/0c8de3fca4a702f8ff5c435e67f378d1fce70243/models/tf.py#L307-L309
            # See this PR for details: https://github.com/ultralytics/ultralytics/pull/1695
            img_h = shape[2] * self.stride[0]
            img_w = shape[3] * self.stride[0]
            img_size = torch.tensor([img_w, img_h, img_w, img_h], device=dbox.device).reshape(1, 4, 1)
            dbox /= img_size

        y = torch.cat((dbox, cls.sigmoid()), 1)
        return y if self.export else (y, x)

    def bias_init(self):
        """Initialize Detect() biases, WARNING: requires stride availability."""
        m = self  # self.model[-1]  # Detect() module
        # cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1
        # ncf = math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum())  # nominal class frequency
        for a, b, s in zip(m.cv2, m.cv3, m.stride):  # from
            a[-1].bias.data[:] = 1.0  # box
            b[-1].bias.data[:m.nc] = math.log(5 / m.nc / (640 / s) ** 2)  # cls (.01 objects, 80 classes, 640 img)

实验

脚本

import warnings
warnings.filterwarnings('ignore')
from ultralytics import YOLO

if __name__ == '__main__':
#     修改为自己的配置文件地址
    model = YOLO('/root/ultralytics-main/ultralytics/cfg/models/11/yolov11-Detect_DyHead.yaml')
#     修改为自己的数据集地址
    model.train(data='/root/ultralytics-main/ultralytics/cfg/datasets/coco8.yaml',
                cache=False,
                imgsz=640,
                epochs=10,
                single_cls=False,  # 是否是单类别检测
                batch=8,
                close_mosaic=10,
                workers=0,
                optimizer='SGD',
                amp=True,
                project='runs/train',
                name='Detect_DyHead',
                )

结果