【深入浅出PyTorch】5-模型定义

5-模型定义

5.1-模型定义方式

5.1.1-Sequential

直接顺序:

## Sequential: Direct list
# 直接顺序
import torch.nn as nn
net1 = nn.Sequential(
        nn.Linear(784, 256),
        nn.ReLU(),
        nn.Linear(256, 10), 
        )
print(net1)

有序字典

## Sequential: Ordered Dict
# 有序字典，可以实现模型名称
import collections
import torch.nn as nn
net2 = nn.Sequential(collections.OrderedDict([
          ('fc1', nn.Linear(784, 256)),
          ('relu1', nn.ReLU()),
          ('fc2', nn.Linear(256, 10))
          ]))
print(net2)

实际上两种方式并没有质的区别

# 试一下
# 检验两种方式输出的结果是否相同
a = torch.rand(4,784)
out1 = net1(a)
out2 = net2(a)
print(out1.shape==out2.shape, out1.shape)
# True torch.Size([4, 10])

5.1.2-ModuleList

## ModuleList
# 创建一个模型List
net3 = nn.ModuleList([nn.Linear(784, 256), nn.ReLU()])
# 向模型List中添加
net3.append(nn.Linear(256, 10)) # # 类似List的append操作
print(net3[-1])  # 类似List的索引访问
print(net3)

定义后没有形成网络，需要自行实现

class Net3(nn.Module):
    def __init__(self):
        super().__init__()
        self.modulelist = nn.ModuleList([nn.Linear(784, 256), nn.ReLU()])
        self.modulelist.append(nn.Linear(256, 10))
    
    def forward(self, x):
        for layer in self.modulelist:
            x = layer(x)
        return x
net3_ = Net3()
out3_ = net3_(a)
print(out3_.shape)

5.1.3-ModuleDict

类似Dict类型的Sequential和ModuleList的结合体，没有形成网络

net4 = nn.ModuleDict({
    'linear': nn.Linear(784, 256),
    'act': nn.ReLU(),
})
net4['output'] = nn.Linear(256, 10) # 添加
print(net4['linear']) # 访问
print(net4.output)

5.2-利用模块快速搭建网络

5.2.1-U-Net

U-Net是分割 (Segmentation) 模型的杰作，在以医学影像为代表的诸多领域有着广泛的应用。U-Net模型结构如下图所示，通过残差连接结构解决了模型学习中的退化问题，使得神经网络的深度能够不断扩展。

结合上图，不难发现U-Net模型具有非常好的对称性。模型从上到下分为若干层，每层由左侧和右侧两个模型块组成，每侧的模型块与其上下模型块之间有连接；同时位于同一层左右两侧的模型块之间也有连接，称为“Skip-connection”。此外还有输入和输出处理等其他组成部分。由于模型的形状非常像英文字母的“U”，因此被命名为“U-Net”。

组成U-Net的模型块主要有如下几个部分：

1. 每个子块内部的两次卷积（Double Convolution）
2. 左侧模型块之间的下采样连接，即最大池化（Max pooling）
3. 右侧模型块之间的上采样连接（Up sampling）
4. 输出层的处理

5.2.2-定义模块

import os
import numpy as np
import collections
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchvision

首先定义几个模块：

class DoubleConv(nn.Module):
    """(convolution => [BN] => ReLU) * 2"""

    def __init__(self, in_channels, out_channels, mid_channels=None):
        super().__init__()
        if not mid_channels:
            mid_channels = out_channels
        self.double_conv = nn.Sequential(
            nn.Conv2d(in_channels, mid_channels, kernel_size=3, padding=1, bias=False),
            nn.BatchNorm2d(mid_channels),
            nn.ReLU(inplace=True),
            nn.Conv2d(mid_channels, out_channels, kernel_size=3, padding=1, bias=False),
            nn.BatchNorm2d(out_channels),
            nn.ReLU(inplace=True)
        )

    def forward(self, x):
        return self.double_conv(x)

class Down(nn.Module):
    """Downscaling with maxpool then double conv"""

    def __init__(self, in_channels, out_channels):
        super().__init__()
        self.maxpool_conv = nn.Sequential(
            nn.MaxPool2d(2),
            DoubleConv(in_channels, out_channels)
        )

    def forward(self, x):
        return self.maxpool_conv(x)

class Up(nn.Module):
    """Upscaling then double conv"""

    def __init__(self, in_channels, out_channels, bilinear=True):
        super().__init__()

        # if bilinear, use the normal convolutions to reduce the number of channels
        if bilinear:
            self.up = nn.Upsample(scale_factor=2, mode='bilinear', align_corners=True)
            self.conv = DoubleConv(in_channels, out_channels, in_channels // 2)
        else:
            self.up = nn.ConvTranspose2d(in_channels, in_channels // 2, kernel_size=2, stride=2)
            self.conv = DoubleConv(in_channels, out_channels)

    def forward(self, x1, x2):
        x1 = self.up(x1)
        # input is CHW
        diffY = x2.size()[2] - x1.size()[2]
        diffX = x2.size()[3] - x1.size()[3]

        x1 = F.pad(x1, [diffX // 2, diffX - diffX // 2,
                        diffY // 2, diffY - diffY // 2])
        # if you have padding issues, see
        # https://github.com/HaiyongJiang/U-Net-Pytorch-Unstructured-Buggy/commit/0e854509c2cea854e247a9c615f175f76fbb2e3a
        # https://github.com/xiaopeng-liao/Pytorch-UNet/commit/8ebac70e633bac59fc22bb5195e513d5832fb3bd
        x = torch.cat([x2, x1], dim=1)
        return self.conv(x)

class OutConv(nn.Module):
    def __init__(self, in_channels, out_channels):
        super(OutConv, self).__init__()
        self.conv = nn.Conv2d(in_channels, out_channels, kernel_size=1)

    def forward(self, x):
        return self.conv(x)

5.2.3-利用模型块组装U-Net

然后使用UNet组装网络

## 组装
class UNet(nn.Module):
    def __init__(self, n_channels, n_classes, bilinear=True):
        super(UNet, self).__init__()
        self.n_channels = n_channels
        self.n_classes = n_classes
        self.bilinear = bilinear
		# 在这里组装网络
        self.inc = DoubleConv(n_channels, 64)
        self.down1 = Down(64, 128)
        self.down2 = Down(128, 256)
        self.down3 = Down(256, 512)
        factor = 2 if bilinear else 1
        self.down4 = Down(512, 1024 // factor)
        self.up1 = Up(1024, 512 // factor, bilinear)
        self.up2 = Up(512, 256 // factor, bilinear)
        self.up3 = Up(256, 128 // factor, bilinear)
        self.up4 = Up(128, 64, bilinear)
        self.outc = OutConv(64, n_classes)

    def forward(self, x):
        x1 = self.inc(x)
        x2 = self.down1(x1)
        x3 = self.down2(x2)
        x4 = self.down3(x3)
        x5 = self.down4(x4)
        x = self.up1(x5, x4)
        x = self.up2(x, x3)
        x = self.up3(x, x2)
        x = self.up4(x, x1)
        logits = self.outc(x)
        return logits