pytorch從頭開始搭建UNet++的過(guò)程詳解

更新時(shí)間：2022年11月01日 10:46:35 作者：楚楚小甜心

大家都知道Unet是一個(gè)最近比較火的網(wǎng)絡(luò)結(jié)構(gòu)，這篇文章主要介紹了pytorch從頭開始搭建UNet++的過(guò)程詳解,需要的朋友可以參考下

Unet++代碼

網(wǎng)絡(luò)架構(gòu)

黑色部分是Backbone，是原先的UNet。

綠色箭頭為上采樣，藍(lán)色箭頭為密集跳躍連接。

綠色的模塊為密集連接塊，是經(jīng)過(guò)左邊兩個(gè)部分拼接操作后組成的

Backbone

2個(gè)3x3的卷積，padding=1。

class VGGBlock(nn.Module):
    def __init__(self, in_channels, middle_channels, out_channels):
        super().__init__()
        self.relu = nn.ReLU(inplace=True)
        self.conv1 = nn.Conv2d(in_channels, middle_channels, 3, padding=1)
        self.bn1 = nn.BatchNorm2d(middle_channels)
        self.conv2 = nn.Conv2d(middle_channels, out_channels, 3, padding=1)
        self.bn2 = nn.BatchNorm2d(out_channels)

    def forward(self, x):
        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)
        out = self.relu(out)

        return out

上采樣

圖中的綠色箭頭，上采樣使用雙線性插值。

雙線性插值就是有兩個(gè)變量的插值函數(shù)的線性插值擴(kuò)展，其核心思想是在兩個(gè)方向分別進(jìn)行一次線性插值

torch.nn.Upsample(size=None, scale_factor=None, mode='nearest', align_corners=None, recompute_scale_factor=None)

參數(shù)說(shuō)明：
①size：可以用來(lái)指定輸出空間的大小，默認(rèn)是None；
②scale_factor：比例因子，比如scale_factor=2意味著將輸入圖像上采樣2倍，默認(rèn)是None；
③mode：用來(lái)指定上采樣算法，有’nearest’、 ‘linear’、‘bilinear’、‘bicubic’、‘trilinear’，默認(rèn)是’nearest’。上采樣算法在本文中會(huì)有詳細(xì)理論進(jìn)行講解；
④align_corners：如果True，輸入和輸出張量的角像素對(duì)齊，從而保留這些像素的值，默認(rèn)是False。此處True和False的區(qū)別本文中會(huì)有詳細(xì)的理論講解；
⑤recompute_scale_factor：如果recompute_scale_factor是True，則必須傳入scale_factor并且scale_factor用于計(jì)算輸出大小。計(jì)算出的輸出大小將用于推斷插值的新比例。請(qǐng)注意，當(dāng)scale_factor為浮點(diǎn)數(shù)時(shí)，由于舍入和精度問(wèn)題，它可能與重新計(jì)算的scale_factor不同。如果recompute_scale_factor是False，那么size或scale_factor將直接用于插值。

class Up(nn.Module):
    def __init__(self):
        super().__init__()
        self.up = nn.Upsample(scale_factor=2, mode='bilinear', align_corners=True)

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

        x1 = F.pad(x1, [diffX // 2, diffX - diffX // 2,
                        diffY // 2, diffY - diffY // 2])
        x = torch.cat([x2, x1], dim=1)
        return x

下采樣

圖中的黑色箭頭，采用的是最大池化。

self.pool = nn.MaxPool2d(2, 2)

深度監(jiān)督

所示，該結(jié)構(gòu)下有4個(gè)分支，可以分為兩種模式。

精確模式：4個(gè)分支取平均值結(jié)果

快速模式：只選擇一個(gè)分支，其余被剪枝

if self.deep_supervision:
   output1 = self.final1(x0_1)
   output2 = self.final2(x0_2)
   output3 = self.final3(x0_3)
   output4 = self.final4(x0_4)
   return [output1, output2, output3, output4]

else:
    output = self.final(x0_4)
     return output

網(wǎng)絡(luò)架構(gòu)代碼

class NestedUNet(nn.Module):
    def __init__(self, num_classes=1, input_channels=1, deep_supervision=False, **kwargs):
        super().__init__()

        nb_filter = [32, 64, 128, 256, 512]

        self.deep_supervision = deep_supervision

        self.pool = nn.MaxPool2d(2, 2)
        self.up = Up()
  
        self.conv0_0 = VGGBlock(input_channels, nb_filter[0], nb_filter[0])
        self.conv1_0 = VGGBlock(nb_filter[0], nb_filter[1], nb_filter[1])
        self.conv2_0 = VGGBlock(nb_filter[1], nb_filter[2], nb_filter[2])
        self.conv3_0 = VGGBlock(nb_filter[2], nb_filter[3], nb_filter[3])
        self.conv4_0 = VGGBlock(nb_filter[3], nb_filter[4], nb_filter[4])

        self.conv0_1 = VGGBlock(nb_filter[0]+nb_filter[1], nb_filter[0], nb_filter[0])
        self.conv1_1 = VGGBlock(nb_filter[1]+nb_filter[2], nb_filter[1], nb_filter[1])
        self.conv2_1 = VGGBlock(nb_filter[2]+nb_filter[3], nb_filter[2], nb_filter[2])
        self.conv3_1 = VGGBlock(nb_filter[3]+nb_filter[4], nb_filter[3], nb_filter[3])

        self.conv0_2 = VGGBlock(nb_filter[0]*2+nb_filter[1], nb_filter[0], nb_filter[0])
        self.conv1_2 = VGGBlock(nb_filter[1]*2+nb_filter[2], nb_filter[1], nb_filter[1])
        self.conv2_2 = VGGBlock(nb_filter[2]*2+nb_filter[3], nb_filter[2], nb_filter[2])

        self.conv0_3 = VGGBlock(nb_filter[0]*3+nb_filter[1], nb_filter[0], nb_filter[0])
        self.conv1_3 = VGGBlock(nb_filter[1]*3+nb_filter[2], nb_filter[1], nb_filter[1])

        self.conv0_4 = VGGBlock(nb_filter[0]*4+nb_filter[1], nb_filter[0], nb_filter[0])

        if self.deep_supervision:
            self.final1 = nn.Conv2d(nb_filter[0], num_classes, kernel_size=1)
            self.final2 = nn.Conv2d(nb_filter[0], num_classes, kernel_size=1)
            self.final3 = nn.Conv2d(nb_filter[0], num_classes, kernel_size=1)
            self.final4 = nn.Conv2d(nb_filter[0], num_classes, kernel_size=1)
        else:
            self.final = nn.Conv2d(nb_filter[0], num_classes, kernel_size=1)
    def forward(self, input):
        x0_0 = self.conv0_0(input)
        x1_0 = self.conv1_0(self.pool(x0_0))
        x0_1 = self.conv0_1(self.up(x1_0, x0_0))

        x2_0 = self.conv2_0(self.pool(x1_0))
        x1_1 = self.conv1_1(self.up(x2_0, x1_0))
        x0_2 = self.conv0_2(self.up(x1_1, torch.cat([x0_0, x0_1], 1)))

        x3_0 = self.conv3_0(self.pool(x2_0))
        x2_1 = self.conv2_1(self.up(x3_0, x2_0))   
        x1_2 = self.conv1_2(self.up(x2_1, torch.cat([x1_0, x1_1], 1)))
        x0_3 = self.conv0_3(self.up(x1_2, torch.cat([x0_0, x0_1, x0_2], 1)))

        x4_0 = self.conv4_0(self.pool(x3_0))
        x3_1 = self.conv3_1(self.up(x4_0, x3_0))
        x2_2 = self.conv2_2(self.up(x3_1, torch.cat([x2_0, x2_1], 1)))
        x1_3 = self.conv1_3(self.up(x2_2, torch.cat([x1_0, x1_1, x1_2], 1)))
        x0_4 = self.conv0_4(self.up(x1_3, torch.cat([x0_0, x0_1, x0_2, x0_3], 1)))

        if self.deep_supervision:
            output1 = self.final1(x0_1)
            output2 = self.final2(x0_2)
            output3 = self.final3(x0_3)
            output4 = self.final4(x0_4)
            return [output1, output2, output3, output4]

        else:
            output = self.final(x0_4)
            return output

到此這篇關(guān)于pytorch從頭開始搭建UNet++的過(guò)程詳解的文章就介紹到這了,更多相關(guān)pytorch搭建UNet++內(nèi)容請(qǐng)搜索腳本之家以前的文章或繼續(xù)瀏覽下面的相關(guān)文章希望大家以后多多支持腳本之家！

您可能感興趣的文章: