YOLOv10全网最新创新点改进系列:YOLOv10改进之添加注意力-ContextAggregation,有效涨点!!!
所有改进代码均经过实验测试跑通!截止发稿时YOLOv10已改进40+!自己排列组合2-4种后,考虑位置不同后可排列组合上千万种!改进不重样!!专注AI学术,关注B站up主:Ai学术叫叫兽er!
购买相关资料 畅享一对一答疑!
# YOLOv10全网最新创新点改进系列:YOLOv10改进之添加注意力-ContextAggregation,有效涨点!!!
一、更改 yaml文件
二、新建ContextAggregation.py
三、更改 tasks.py
详细改进流程和操作,请关注B站博主:AI学术叫叫兽
相关源码已在B站:AI学术叫叫兽
上架!!!!科研搞起来!表情包
论文地址在这
AI学术叫叫兽er在这!家人们,给我遥遥领先!!!
一、更改yaml文件
新建 ContextAggregation.py
#详细改进流程和操作,请关注B站博主:AI学术叫叫兽 持续更新哦
import torch
import torch.nn as nn
from mmcv.cnn import ConvModule, caffe2_xavier_init, constant_init
#详细改进流程和操作,请关注B站博主:AI学术叫叫兽
from mmcv.cnn import ConvModule
#详细改进流程和操作,请关注B站博主:AI学术叫叫兽
class ContextAggregation(nn.Module):
#详细改进流程和操作,请关注B站博主:AI学术叫叫兽
def __init__(self, in_channels, reduction=1, conv_cfg=None):
super(ContextAggregation, self).__init__()
self.in_channels = in_channels
self.reduction = reduction
self.inter_channels = max(in_channels // reduction, 1)
conv_params = dict(kernel_size=1, conv_cfg=conv_cfg, act_cfg=None)
self.a = ConvModule(in_channels, 1, **conv_params)
self.k = ConvModule(in_channels, 1, **conv_params)
self.v = ConvModule(in_channels, self.inter_channels, **conv_params)
self.m = ConvModule(self.inter_channels, in_channels, **conv_params)
self.init_weights()
def init_weights(self):
for m in (self.a, self.k, self.v):
caffe2_xavier_init(m.conv)
constant_init(self.m.conv, 0)
def forward(self, x):
n, c = x.size(0), self.inter_channels
# a: [N, 1, H, W]
a = self.a(x).sigmoid()
# k: [N, 1, HW, 1]
k = self.k(x).view(n, 1, -1, 1).softmax(2)
# v: [N, 1, C, HW]
v = self.v(x).view(n, 1, c, -1)
# y: [N, C, 1, 1]
y = torch.matmul(v, k).view(n, c, 1, 1)
y = self.m(y) * a
return x + y
#详细改进流程和操作,请关注B站博主:AI学术叫叫兽 片
更改 tasks.py
找到tasks.py中的此代码,替换即可,大约在650行左右。
if m in (Classify, Conv, GGhostRegNet, ConvTranspose, GhostConv, Bottleneck, GhostBottleneck, SPP, SPPF, DWConv, Focus,
BottleneckCSP, C1, C2, C2f, C3, C3TR, C3Ghost, nn.ConvTranspose2d, DWConvTranspose2d, C3x, RepC3, SEAttention,ContextAggregation):
c1, c2 = ch[f], args[0]
if c2 != nc: # if c2 not equal to number of classes (i.e. for Classify() output)
c2 = make_divisible(min(c2, max_channels) * width, 8)
检查修改是否成功
注意!
别忘喽~关注B站博主:AI学术叫叫兽
往期B站视频已经更新了四层检测层,如果注意力加四个检测头,会发生什么?快动手去试试!
科研搞起来!一Giao窝里Giao Giao!!
已经更新了 注意力、特征提取网络、添加检测头、优化卷积操作等改进方法。
改进方法持续更新,应B站粉丝要求,近期会在B站开设论文写作方面的专栏。