摘要：涨点神器！利用Involution(内卷)，可构建用于视觉识别的新一代神经网络！在分类、检测和分割任务上涨点显著！
1.Involution(内卷)
涨点神器！利用Involution(内卷)，可构建用于视觉识别的新一代神经网络！在分类、检测和分割任务上涨点显著！
Inverting the Inherence of Convolution for Visual Recognition（CVPR2021）
论文链接：https://arxiv.org/abs/2103.06255 github代码链接：https://github.com/d-li14/involution
作者认为卷积操作的两个特征虽然也有一定的优势，但同样也有缺点。所以提出了Involution，Involution所拥有的特征正好和卷积相对称，即 spatial-specific and channel-agnostic
那就是通道无关和特定于空间。和卷积一样，内卷也有内卷核（involution kernels）。内卷核在空间范围上是不同的，但在通道之间共享。看到这里就有一定的画面感了。
内卷的优点：
1.可以在更大的空间范围中总结上下文信息，从而克服long-range interaction（本来的卷积操作只能在特定的小空间如3x3中集合空间信息）
2.内卷可以将权重自适应地分配到不同的位置，从而对空间域中信息量最大的视觉元素进行优先级排序。（本来的卷积在空间的每一个地方都是用到同一个卷积核，用的同一套权重）
2.Yolov5加入Involution
2.1 Involution加入common.py中：
代码语言：javascript复制class Involution(nn.Module):
def __init__(self,c1,c2,kernel_size,stride):
super(Involution, self).__init__()
self.kernel_size = kernel_size
self.stride = stride
self.c1 = c1
reduction_ratio = 4
self.group_channels = 16
self.groups = self.c1 // self.group_channels
self.conv1 = Conv(
c1, c1 // reduction_ratio,1)
self.conv2 = Conv(
c1 // reduction_ratio,
kernel_size**2 * self.groups,
1,1)
if stride > 1:
self.avgpool = nn.AvgPool2d(stride, stride)
self.unfold = nn.Unfold(kernel_size, 1, (kernel_size-1)//2, stride)
def forward(self, x):
weight = self.conv2(self.conv1(x if self.stride == 1 else self.avgpool(x)))
b, c, h, w = weight.shape
weight = weight.view(b, self.groups, self.kernel_size**2, h, w).unsqueeze(2)
out = self.unfold(x).view(b, self.groups, self.group_channels, self.kernel_size**2, h, w)
out = (weight * out).sum(dim=3).view(b, self.c1, h, w)
return out2.2 修改yolov5s_involution.yaml
代码语言：javascript复制# parameters
nc: 1 # number of classes
depth_multiple: 0.33 # model depth multiple
width_multiple: 0.50 # layer channel multiple
# anchors
anchors:
- [10,13, 16,30, 33,23] # P3/8
- [30,61, 62,45, 59,119] # P4/16
- [116,90, 156,198, 373,326] # P5/32
# YOLOv5 backbone
backbone:
# [from, number, module, args]
[[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2
[-1, 1, Conv, [128, 3, 2]],
[-1, 1, Involution, [128,3,2]], # 2-P2/4
[-1, 3, C3, [128,True]],
[-1, 1, Conv, [256, 3, 2]],
[-1, 1, Involution, [256,3,2]], # 5-P3/8
[-1, 6, C3, [256,True]],
[-1, 1, Conv, [512,3,2]], #7
[-1, 1, Involution, [512,3,1]], # 8-P4/16
[-1, 9, C3, [512]],
[-1, 1, Conv, [1024, 3, 2]],
[-1, 1, Involution, [1024,3,1]], # 11-P5/32
[-1, 3, C3, [1024]],
[-1, 1, SPPF, [1024,5]],
]
# YOLOv5 head
head:
[[-1, 1, Conv, [512, 1, 1]],
[-1, 1, nn.Upsample, [None, 2, 'nearest']],
[[-1, 8], 1, Concat, [1]], # cat backbone P4
[-1, 3, C3, [512, False]], # 17
[-1, 1, Conv, [256, 1, 1]],
[-1, 1, nn.Upsample, [None, 2, 'nearest']],
[[-1, 5], 1, Concat, [1]], # cat backbone P3
[-1, 3, C3, [256, False]], # 21 (P3/8-small)
[-1, 1, Conv, [256, 3, 2]],
[[-1, 18], 1, Concat, [1]], # cat head P4
[-1, 3, C3, [512, False]], # 24 (P4/16-medium)
[-1, 1, Conv, [512, 3, 2]],
[[-1, 14], 1, Concat, [1]], # cat head P5
[-1, 3, C3, [1024, False]], # 27 (P5/32-large)
[[21, 24, 27], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5)
]
by CSDN AI小怪兽 https://cv2023.blog.csdn.net/article/details/129621812
我正在参与2023腾讯技术创作特训营第三期有奖征文，组队打卡瓜分大奖！
