【26】unet网络复现及学习（总述）

【1】模型介绍

---

    Layer (type)               Output Shape         Param #

================================================================
Conv2d-1 [-1, 64, 512, 512] 640
BatchNorm2d-2 [-1, 64, 512, 512] 128
ReLU-3 [-1, 64, 512, 512] 0
Conv2d-4 [-1, 64, 512, 512] 36,928
BatchNorm2d-5 [-1, 64, 512, 512] 128
ReLU-6 [-1, 64, 512, 512] 0
DoubleConv-7 [-1, 64, 512, 512] 0
MaxPool2d-8 [-1, 64, 256, 256] 0
Conv2d-9 [-1, 128, 256, 256] 73,856
BatchNorm2d-10 [-1, 128, 256, 256] 256
ReLU-11 [-1, 128, 256, 256] 0
Conv2d-12 [-1, 128, 256, 256] 147,584
BatchNorm2d-13 [-1, 128, 256, 256] 256
ReLU-14 [-1, 128, 256, 256] 0
DoubleConv-15 [-1, 128, 256, 256] 0
MaxPool2d-16 [-1, 128, 128, 128] 0
Conv2d-17 [-1, 256, 128, 128] 295,168
BatchNorm2d-18 [-1, 256, 128, 128] 512
ReLU-19 [-1, 256, 128, 128] 0
Conv2d-20 [-1, 256, 128, 128] 590,080
BatchNorm2d-21 [-1, 256, 128, 128] 512
ReLU-22 [-1, 256, 128, 128] 0
DoubleConv-23 [-1, 256, 128, 128] 0
MaxPool2d-24 [-1, 256, 64, 64] 0
Conv2d-25 [-1, 512, 64, 64] 1,180,160
BatchNorm2d-26 [-1, 512, 64, 64] 1,024
ReLU-27 [-1, 512, 64, 64] 0
Conv2d-28 [-1, 512, 64, 64] 2,359,808
BatchNorm2d-29 [-1, 512, 64, 64] 1,024
ReLU-30 [-1, 512, 64, 64] 0
DoubleConv-31 [-1, 512, 64, 64] 0
MaxPool2d-32 [-1, 512, 32, 32] 0
Conv2d-33 [-1, 1024, 32, 32] 4,719,616
BatchNorm2d-34 [-1, 1024, 32, 32] 2,048
ReLU-35 [-1, 1024, 32, 32] 0
Conv2d-36 [-1, 1024, 32, 32] 9,438,208
BatchNorm2d-37 [-1, 1024, 32, 32] 2,048
ReLU-38 [-1, 1024, 32, 32] 0
DoubleConv-39 [-1, 1024, 32, 32] 0
ConvTranspose2d-40 [-1, 512, 64, 64] 2,097,664
Conv2d-41 [-1, 512, 64, 64] 4,719,104
BatchNorm2d-42 [-1, 512, 64, 64] 1,024
ReLU-43 [-1, 512, 64, 64] 0
Conv2d-44 [-1, 512, 64, 64] 2,359,808
BatchNorm2d-45 [-1, 512, 64, 64] 1,024
ReLU-46 [-1, 512, 64, 64] 0
DoubleConv-47 [-1, 512, 64, 64] 0
ConvTranspose2d-48 [-1, 256, 128, 128] 524,544
Conv2d-49 [-1, 256, 128, 128] 1,179,904
BatchNorm2d-50 [-1, 256, 128, 128] 512
ReLU-51 [-1, 256, 128, 128] 0
Conv2d-52 [-1, 256, 128, 128] 590,080
BatchNorm2d-53 [-1, 256, 128, 128] 512
ReLU-54 [-1, 256, 128, 128] 0
DoubleConv-55 [-1, 256, 128, 128] 0
ConvTranspose2d-56 [-1, 128, 256, 256] 131,200
Conv2d-57 [-1, 128, 256, 256] 295,040
BatchNorm2d-58 [-1, 128, 256, 256] 256
ReLU-59 [-1, 128, 256, 256] 0
Conv2d-60 [-1, 128, 256, 256] 147,584
BatchNorm2d-61 [-1, 128, 256, 256] 256
ReLU-62 [-1, 128, 256, 256] 0
DoubleConv-63 [-1, 128, 256, 256] 0
ConvTranspose2d-64 [-1, 64, 512, 512] 32,832
Conv2d-65 [-1, 64, 512, 512] 73,792
BatchNorm2d-66 [-1, 64, 512, 512] 128
ReLU-67 [-1, 64, 512, 512] 0
Conv2d-68 [-1, 64, 512, 512] 36,928
BatchNorm2d-69 [-1, 64, 512, 512] 128
ReLU-70 [-1, 64, 512, 512] 0
DoubleConv-71 [-1, 64, 512, 512] 0
Conv2d-72 [-1, 2, 512, 512] 130
Total params: 31,042,434
Trainable params: 31,042,434
Non-trainable params: 0

【2】文件夹介绍

unet
│ config.ini #配置文件
│ Digraph.gv
│ main.py #主要程序、训练、测试函数
│ Read.md #unet网络介绍
│ setdata.py #图像数据设置和读取
│ structure.txt #文件树
│ testnumpy.py #函数测试文件
│ Unetmodel.py #unet模型文件
│ visualize.py #可视化辅助文件
│
├─.vscode
│ settings.json
│
├─dowoload #权重文件
│ weights_19.pth
│
├─modelimage #模型可视化结果
│ Digraph.gv.pdf
│ unet_model_image.png
│
└─__pycache__
setdata.cpython-36.pyc
Unetmodel.cpython-36.pyc
visualize.cpython-36.pyc