摘要:SKAttention注意力助力YOLOv5,即插即用,性能优于SENet
1. SKAttention
论文:https://arxiv.org/pdf/1903.06586.pdf
多个 SK 块的堆叠得到 SKNet,这个名字也是为了致敬 SENet。
SKNet 在 ImageNet、CIFAR 数据集上都取得了 SOTA。
详细的实验分析表明,SKNet 中的神经元可以捕获具有不同比例的目标对象,实验验证了神经元根据输入自适应地调整其感受野大小的能力。
本文的方法分为三个部分:Split,Fuse,Select。Split就是一个multi-branch的操作,用不同的卷积核进行卷积得到不同的特征;Fuse部分就是用SE的结构获取通道注意力的矩阵(N个卷积核就可以得到N个注意力矩阵,这步操作对所有的特征参数共享),这样就可以得到不同kernel经过SE之后的特征;Select操作就是将这几个特征进行相加。
1.1 加入 common.py
中
代码语言:javascript复制###################### SKAttention #### start by AI&CV ###############################
from torch.nn import init
from collections import OrderedDict
class SKAttention(nn.Module):
def __init__(self, channel=512, kernels=[1, 3, 5, 7], reduction=16, group=1, L=32):
super().__init__()
self.d = max(L, channel // reduction)
self.convs = nn.ModuleList([])
for k in kernels:
self.convs.append(
nn.Sequential(OrderedDict([
('conv', nn.Conv2d(channel, channel, kernel_size=k, padding=k // 2, groups=group)),
('bn', nn.BatchNorm2d(channel)),
('relu', nn.ReLU())
]))
)
self.fc = nn.Linear(channel, self.d)
self.fcs = nn.ModuleList([])
for i in range(len(kernels)):
self.fcs.append(nn.Linear(self.d, channel))
self.softmax = nn.Softmax(dim=0)
def forward(self, x):
bs, c, _, _ = x.size()
conv_outs = []
### split
for conv in self.convs:
conv_outs.append(conv(x))
feats = torch.stack(conv_outs, 0) # k,bs,channel,h,w
### fuse
U = sum(conv_outs) # bs,c,h,w
### reduction channel
S = U.mean(-1).mean(-1) # bs,c
Z = self.fc(S) # bs,d
### calculate attention weight
weights = []
for fc in self.fcs:
weight = fc(Z)
weights.append(weight.view(bs, c, 1, 1)) # bs,channel
attention_weughts = torch.stack(weights, 0) # k,bs,channel,1,1
attention_weughts = self.softmax(attention_weughts) # k,bs,channel,1,1
### fuse
V = (attention_weughts * feats).sum(0)
return V
###################### SKAttention #### end by AI&CV ###############################
1.2 加入yolo
.py中:
代码语言:javascript复制 elif m is SKAttention:
c1, c2 = ch[f], args[0]
if c2 != nc:
c2 = make_divisible(min(c2, max_channels) * width, 8)
args = [c1, *args[1:]]
1.3 yolov5s_SKAttention.yaml
代码语言:javascript复制# YOLOv5