【Unet系列】
https://tianfeng.space/1947.html
前言概念
图像分割
分割任务就是在原始图像中逐像素的找到你需要的家伙!
语义分割
就是把每个像素都打上标签（这个像素点是人，树，背景等）
（语义分割只区分类别，不区分类别中具体单位）
实例分割
实例分割不光要区别类别，还要区分类别中每一个个体
损失函数：
逐像素的交叉熵：还经常需要考虑样本均衡问题，交叉熵损失函数公式如下：
Focal loss：样本也由难易之分，就跟玩游戏一样，难度越高的BOSS奖励越高
Gamma通常设置为2，例如预测正样本概率0.95，如果预测正样本概率0.4， （相当于样本的难易权值）
（再结合样本数量的权值就是Focal Loss）
IOU计算
多分类任务时：iou_dog = 801 / true_dog + predict_dog - 801
MIOU指标：
MIOU就是计算所有类别的平均值，一般当作分割任务评估指标
Unet
整体结构：概述就是编码解码过程；简单但是很实用，应用广；起初是做医学方向，现在也是
Unet++
整体网络结构：特征融合，拼接更全面；其实跟densenet思想一致；把能拼能凑的特征全用上
Deep Supervision ：多输出损失；由多个位置计算，再更新
容易剪枝：可以根据速度要求来快速完成剪枝；训练的时候同样会用到L4，效果还不错
U²net
代码 论文
听名字知道就是把Unet中每个stage再变成一个Unet，这样就嵌套了一个Unet变成U²net；
输出为解码器各个阶段输出再拼接，经过一次卷积输出
现有卷积块和我们提出的残差U形块RSU的说明：（a）普通卷积块PLN，（b）残差类块RES，（c）密集类块DSE，（d）启始类块INC和（e）我们的残差U型块RSU
残差块与我们的RSU的比较
就作者展示的效果而言，出奇的不错，有兴趣去代码界面看看，使用也很简单，下面展示一些
代码结构放最后；有兴趣看看
#U²net结构；387行forward开始
import torch
import torch.nn as nn
from torchvision import models
import torch.nn.functional as F
class REBNCONV(nn.Module):
def __init__(self,in_ch=3,out_ch=3,dirate=1):
super(REBNCONV,self).__init__()
self.conv_s1 = nn.Conv2d(in_ch,out_ch,3,padding=1*dirate,dilation=1*dirate)
self.bn_s1 = nn.BatchNorm2d(out_ch)
self.relu_s1 = nn.ReLU(inplace=True)
def forward(self,x):
hx = x
xout = self.relu_s1(self.bn_s1(self.conv_s1(hx)))
return xout
## upsample tensor 'src' to have the same spatial size with tensor 'tar'
def _upsample_like(src,tar):
src = F.upsample(src,size=tar.shape[2:],mode='bilinear')
return src
### RSU-7 ###
class RSU7(nn.Module):#UNet07DRES(nn.Module):
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
super(RSU7,self).__init__()
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
self.pool1 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool2 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool3 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool4 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv5 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool5 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv6 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.rebnconv7 = REBNCONV(mid_ch,mid_ch,dirate=2)
self.rebnconv6d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv5d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv4d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
def forward(self,x):
print(x.shape)
hx = x
hxin = self.rebnconvin(hx)
print(hxin.shape)
hx1 = self.rebnconv1(hxin)
print(hx1.shape)
hx = self.pool1(hx1)
print(hx.shape)
hx2 = self.rebnconv2(hx)
print(hx2.shape)
hx = self.pool2(hx2)
print(hx.shape)
hx3 = self.rebnconv3(hx)
print(hx3.shape)
hx = self.pool3(hx3)
print(hx.shape)
hx4 = self.rebnconv4(hx)
print(hx4.shape)
hx = self.pool4(hx4)
print(hx.shape)
hx5 = self.rebnconv5(hx)
print(hx5.shape)
hx = self.pool5(hx5)
print(hx.shape)
hx6 = self.rebnconv6(hx)
print(hx6.shape)
hx7 = self.rebnconv7(hx6)
print(hx7.shape)
hx6d = self.rebnconv6d(torch.cat((hx7,hx6),1))
print(hx6d.shape)
hx6dup = _upsample_like(hx6d,hx5)
print(hx6dup.shape)
hx5d = self.rebnconv5d(torch.cat((hx6dup,hx5),1))
print(hx5d.shape)
hx5dup = _upsample_like(hx5d,hx4)
print(hx5dup.shape)
hx4d = self.rebnconv4d(torch.cat((hx5dup,hx4),1))
print(hx4d.shape)
hx4dup = _upsample_like(hx4d,hx3)
print(hx4dup.shape)
hx3d = self.rebnconv3d(torch.cat((hx4dup,hx3),1))
print(hx3d.shape)
hx3dup = _upsample_like(hx3d,hx2)
print(hx3dup.shape)
hx2d = self.rebnconv2d(torch.cat((hx3dup,hx2),1))
print(hx2d.shape)
hx2dup = _upsample_like(hx2d,hx1)
print(hx2dup.shape)
hx1d = self.rebnconv1d(torch.cat((hx2dup,hx1),1))
print(hx1d.shape)
return hx1d + hxin
### RSU-6 ###
class RSU6(nn.Module):#UNet06DRES(nn.Module):
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
super(RSU6,self).__init__()
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
self.pool1 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool2 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool3 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool4 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv5 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.rebnconv6 = REBNCONV(mid_ch,mid_ch,dirate=2)
self.rebnconv5d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv4d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
def forward(self,x):
hx = x
hxin = self.rebnconvin(hx)
hx1 = self.rebnconv1(hxin)
hx = self.pool1(hx1)
hx2 = self.rebnconv2(hx)
hx = self.pool2(hx2)
hx3 = self.rebnconv3(hx)
hx = self.pool3(hx3)
hx4 = self.rebnconv4(hx)
hx = self.pool4(hx4)
hx5 = self.rebnconv5(hx)
hx6 = self.rebnconv6(hx5)
hx5d = self.rebnconv5d(torch.cat((hx6,hx5),1))
hx5dup = _upsample_like(hx5d,hx4)
hx4d = self.rebnconv4d(torch.cat((hx5dup,hx4),1))
hx4dup = _upsample_like(hx4d,hx3)
hx3d = self.rebnconv3d(torch.cat((hx4dup,hx3),1))
hx3dup = _upsample_like(hx3d,hx2)
hx2d = self.rebnconv2d(torch.cat((hx3dup,hx2),1))
hx2dup = _upsample_like(hx2d,hx1)
hx1d = self.rebnconv1d(torch.cat((hx2dup,hx1),1))
return hx1d + hxin
### RSU-5 ###
class RSU5(nn.Module):#UNet05DRES(nn.Module):
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
super(RSU5,self).__init__()
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
self.pool1 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool2 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool3 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.rebnconv5 = REBNCONV(mid_ch,mid_ch,dirate=2)
self.rebnconv4d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
def forward(self,x):
hx = x
hxin = self.rebnconvin(hx)
hx1 = self.rebnconv1(hxin)
hx = self.pool1(hx1)
hx2 = self.rebnconv2(hx)
hx = self.pool2(hx2)
hx3 = self.rebnconv3(hx)
hx = self.pool3(hx3)
hx4 = self.rebnconv4(hx)
hx5 = self.rebnconv5(hx4)
hx4d = self.rebnconv4d(torch.cat((hx5,hx4),1))
hx4dup = _upsample_like(hx4d,hx3)
hx3d = self.rebnconv3d(torch.cat((hx4dup,hx3),1))
hx3dup = _upsample_like(hx3d,hx2)
hx2d = self.rebnconv2d(torch.cat((hx3dup,hx2),1))
hx2dup = _upsample_like(hx2d,hx1)
hx1d = self.rebnconv1d(torch.cat((hx2dup,hx1),1))
return hx1d + hxin
### RSU-4 ###
class RSU4(nn.Module):#UNet04DRES(nn.Module):
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
super(RSU4,self).__init__()
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
self.pool1 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.pool2 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=1)
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=2)
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=1)
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
def forward(self,x):
hx = x
hxin = self.rebnconvin(hx)
hx1 = self.rebnconv1(hxin)
hx = self.pool1(hx1)
hx2 = self.rebnconv2(hx)
hx = self.pool2(hx2)
hx3 = self.rebnconv3(hx)
hx4 = self.rebnconv4(hx3)
hx3d = self.rebnconv3d(torch.cat((hx4,hx3),1))
hx3dup = _upsample_like(hx3d,hx2)
hx2d = self.rebnconv2d(torch.cat((hx3dup,hx2),1))
hx2dup = _upsample_like(hx2d,hx1)
hx1d = self.rebnconv1d(torch.cat((hx2dup,hx1),1))
return hx1d + hxin
### RSU-4F ###
class RSU4F(nn.Module):#UNet04FRES(nn.Module):
def __init__(self, in_ch=3, mid_ch=12, out_ch=3):
super(RSU4F,self).__init__()
self.rebnconvin = REBNCONV(in_ch,out_ch,dirate=1)
self.rebnconv1 = REBNCONV(out_ch,mid_ch,dirate=1)
self.rebnconv2 = REBNCONV(mid_ch,mid_ch,dirate=2)
self.rebnconv3 = REBNCONV(mid_ch,mid_ch,dirate=4)
self.rebnconv4 = REBNCONV(mid_ch,mid_ch,dirate=8)
self.rebnconv3d = REBNCONV(mid_ch*2,mid_ch,dirate=4)
self.rebnconv2d = REBNCONV(mid_ch*2,mid_ch,dirate=2)
self.rebnconv1d = REBNCONV(mid_ch*2,out_ch,dirate=1)
def forward(self,x):
hx = x
hxin = self.rebnconvin(hx)
print(hxin.shape)
hx1 = self.rebnconv1(hxin)
print(hx1.shape)
hx2 = self.rebnconv2(hx1)
print(hx2.shape)
hx3 = self.rebnconv3(hx2)
print(hx3.shape)
hx4 = self.rebnconv4(hx3)
print(hx4.shape)
hx3d = self.rebnconv3d(torch.cat((hx4,hx3),1))
print(hx3d.shape)
hx2d = self.rebnconv2d(torch.cat((hx3d,hx2),1))
print(hx2d.shape)
hx1d = self.rebnconv1d(torch.cat((hx2d,hx1),1))
print(hx1d.shape)
return hx1d + hxin
##### U^2-Net ####
class U2NET(nn.Module):
def __init__(self,in_ch=3,out_ch=1):
super(U2NET,self).__init__()
self.stage1 = RSU7(in_ch,32,64)
self.pool12 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage2 = RSU6(64,32,128)
self.pool23 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage3 = RSU5(128,64,256)
self.pool34 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage4 = RSU4(256,128,512)
self.pool45 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage5 = RSU4F(512,256,512)
self.pool56 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage6 = RSU4F(512,256,512)
# decoder
self.stage5d = RSU4F(1024,256,512)
self.stage4d = RSU4(1024,128,256)
self.stage3d = RSU5(512,64,128)
self.stage2d = RSU6(256,32,64)
self.stage1d = RSU7(128,16,64)
self.side1 = nn.Conv2d(64,out_ch,3,padding=1)
self.side2 = nn.Conv2d(64,out_ch,3,padding=1)
self.side3 = nn.Conv2d(128,out_ch,3,padding=1)
self.side4 = nn.Conv2d(256,out_ch,3,padding=1)
self.side5 = nn.Conv2d(512,out_ch,3,padding=1)
self.side6 = nn.Conv2d(512,out_ch,3,padding=1)
self.outconv = nn.Conv2d(6,out_ch,1)
def forward(self,x):
print(x.shape)
hx = x
#stage 1
hx1 = self.stage1(hx)
print(hx1.shape)
hx = self.pool12(hx1)
print(hx.shape)
#stage 2
hx2 = self.stage2(hx)
print(hx2.shape)
hx = self.pool23(hx2)
print(hx.shape)
#stage 3
hx3 = self.stage3(hx)
print(hx3.shape)
hx = self.pool34(hx3)
print(hx.shape)
#stage 4
hx4 = self.stage4(hx)
print(hx4.shape)
hx = self.pool45(hx4)
print(hx.shape)
#stage 5
hx5 = self.stage5(hx)
print(hx5.shape)
hx = self.pool56(hx5)
print(hx.shape)
#stage 6
hx6 = self.stage6(hx)
print(hx6.shape)
hx6up = _upsample_like(hx6,hx5)
print(hx6up.shape)
#-------------------- decoder --------------------
hx5d = self.stage5d(torch.cat((hx6up,hx5),1))
print(hx5d.shape)
hx5dup = _upsample_like(hx5d,hx4)
print(hx5dup.shape)
hx4d = self.stage4d(torch.cat((hx5dup,hx4),1))
print(hx4d.shape)
hx4dup = _upsample_like(hx4d,hx3)
print(hx4dup.shape)
hx3d = self.stage3d(torch.cat((hx4dup,hx3),1))
print(hx3d.shape)
hx3dup = _upsample_like(hx3d,hx2)
print(hx3dup.shape)
hx2d = self.stage2d(torch.cat((hx3dup,hx2),1))
print(hx2d.shape)
hx2dup = _upsample_like(hx2d,hx1)
print(hx2dup.shape)
hx1d = self.stage1d(torch.cat((hx2dup,hx1),1))
print(hx1d.shape)
#side output
d1 = self.side1(hx1d)
print(d1.shape)
d2 = self.side2(hx2d)
print(d2.shape)
d2 = _upsample_like(d2,d1)
print(d2.shape)
d3 = self.side3(hx3d)
print(d3.shape)
d3 = _upsample_like(d3,d1)
print(d3.shape)
d4 = self.side4(hx4d)
print(d4.shape)
d4 = _upsample_like(d4,d1)
print(d4.shape)
d5 = self.side5(hx5d)
print(d5.shape)
d5 = _upsample_like(d5,d1)
print(d5.shape)
d6 = self.side6(hx6)
print(d6.shape)
d6 = _upsample_like(d6,d1)
print(d6.shape)
d0 = self.outconv(torch.cat((d1,d2,d3,d4,d5,d6),1))
print(d0.shape)
return F.sigmoid(d0), F.sigmoid(d1), F.sigmoid(d2), F.sigmoid(d3), F.sigmoid(d4), F.sigmoid(d5), F.sigmoid(d6)
### U^2-Net small ###
class U2NETP(nn.Module):
def __init__(self,in_ch=3,out_ch=1):
super(U2NETP,self).__init__()
self.stage1 = RSU7(in_ch,16,64)
self.pool12 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage2 = RSU6(64,16,64)
self.pool23 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage3 = RSU5(64,16,64)
self.pool34 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage4 = RSU4(64,16,64)
self.pool45 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage5 = RSU4F(64,16,64)
self.pool56 = nn.MaxPool2d(2,stride=2,ceil_mode=True)
self.stage6 = RSU4F(64,16,64)
# decoder
self.stage5d = RSU4F(128,16,64)
self.stage4d = RSU4(128,16,64)
self.stage3d = RSU5(128,16,64)
self.stage2d = RSU6(128,16,64)
self.stage1d = RSU7(128,16,64)
self.side1 = nn.Conv2d(64,out_ch,3,padding=1)
self.side2 = nn.Conv2d(64,out_ch,3,padding=1)
self.side3 = nn.Conv2d(64,out_ch,3,padding=1)
self.side4 = nn.Conv2d(64,out_ch,3,padding=1)
self.side5 = nn.Conv2d(64,out_ch,3,padding=1)
self.side6 = nn.Conv2d(64,out_ch,3,padding=1)
self.outconv = nn.Conv2d(6,out_ch,1)
def forward(self,x):
hx = x
#stage 1
hx1 = self.stage1(hx)
hx = self.pool12(hx1)
#stage 2
hx2 = self.stage2(hx)
hx = self.pool23(hx2)
#stage 3
hx3 = self.stage3(hx)
hx = self.pool34(hx3)
#stage 4
hx4 = self.stage4(hx)
hx = self.pool45(hx4)
#stage 5
hx5 = self.stage5(hx)
hx = self.pool56(hx5)
#stage 6
hx6 = self.stage6(hx)
hx6up = _upsample_like(hx6,hx5)
#decoder
hx5d = self.stage5d(torch.cat((hx6up,hx5),1))
hx5dup = _upsample_like(hx5d,hx4)
hx4d = self.stage4d(torch.cat((hx5dup,hx4),1))
hx4dup = _upsample_like(hx4d,hx3)
hx3d = self.stage3d(torch.cat((hx4dup,hx3),1))
hx3dup = _upsample_like(hx3d,hx2)
hx2d = self.stage2d(torch.cat((hx3dup,hx2),1))
hx2dup = _upsample_like(hx2d,hx1)
hx1d = self.stage1d(torch.cat((hx2dup,hx1),1))
#side output
d1 = self.side1(hx1d)
d2 = self.side2(hx2d)
d2 = _upsample_like(d2,d1)
d3 = self.side3(hx3d)
d3 = _upsample_like(d3,d1)
d4 = self.side4(hx4d)
d4 = _upsample_like(d4,d1)
d5 = self.side5(hx5d)
d5 = _upsample_like(d5,d1)
d6 = self.side6(hx6)
d6 = _upsample_like(d6,d1)
d0 = self.outconv(torch.cat((d1,d2,d3,d4,d5,d6),1))
return F.sigmoid(d0), F.sigmoid(d1), F.sigmoid(d2), F.sigmoid(d3), F.sigmoid(d4), F.sigmoid(d5), F.sigmoid(d6)