【pytorch】实现条件生成对抗网络 CGAN

生成器网络为Unet

 判别器为patchGAN

代码示例：

class DoubleConv(nn.Module):
    def __init__(self, in_channels, out_channels):
        super(DoubleConv, self).__init__()
        self.conv = nn.Sequential(
            nn.Conv2d(in_channels, out_channels, 3, 1, 1, bias=False),
            nn.BatchNorm2d(out_channels),
            nn.ReLU(inplace=True),
            nn.Conv2d(out_channels, out_channels, 3, 1, 1, bias=False),
            nn.BatchNorm2d(out_channels),
            nn.ReLU(inplace=True),
        )

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

class GeneratorNet(nn.Module):
    def __init__(
            self, in_channels=3, out_channels=1, features=[64, 128, 256, 512],
    ):
        super(GeneratorNet, self).__init__()
        self.ups = nn.ModuleList()
        self.downs = nn.ModuleList()
        self.pool = nn.MaxPool2d(kernel_size=2, stride=2)

        # Down part of UNET :  Refer to pdf file in architecture of generator to understand more about U-net
        for feature in features:
            self.downs.append(DoubleConv(in_channels, feature))
            in_channels = feature

        # Up part of UNET: Refer to pdf file in architecture of generator to understand more about U-net
        for feature in reversed(features):
            self.ups.append(
                nn.ConvTranspose2d(
                    feature*2, feature, kernel_size=2, stride=2,
                )
            )
            self.ups.append(DoubleConv(feature*2, feature))

        self.bottleneck = DoubleConv(features[-1], features[-1]*2)
        self.final_conv = nn.Conv2d(features[0], out_channels, kernel_size=1)

    def forward(self, x):
        skip_connections = []

        for down in self.downs:
            x = down(x)
            skip_connections.append(x)
            x = self.pool(x)

        x = self.bottleneck(x)
        skip_connections = skip_connections[::-1]

        for idx in range(0, len(self.ups), 2):
            x = self.ups[idx](x)
            skip_connection = skip_connections[idx//2]

            if x.shape != skip_connection.shape:
                x = TF.resize(x, size=skip_connection.shape[2:])

            concat_skip = torch.cat((skip_connection, x), dim=1)
            x = self.ups[idx+1](concat_skip)

        return self.final_conv(x)

class DiscriminatorNet(nn.Module):
    def __init__(self, in_channels):
        super().__init__()
        self.disc = nn.Sequential(            
            nn.Conv2d(in_channels, 64, 3, 1, 1), # layer 1 64x256x256
            nn.LeakyReLU(0.01, inplace=True),                     
            nn.MaxPool2d(2),                     # layer 2 64x128x128
            nn.Conv2d(64, 128, 3, 1, 1),         # layer 3 128x128x128
            nn.LeakyReLU(0.01, inplace=True), 
            nn.MaxPool2d(2),                     # layer 4 128x64x64
            nn.Conv2d(128, 256, 3, 1, 1),        # layer 5 256x64x64
            nn.LeakyReLU(0.01, inplace=True),    # batch norm next
            nn.BatchNorm2d(256),
            nn.MaxPool2d(2),                     # layer 6 256x32x32
            nn.Conv2d(256, 256, 3, 1, 1),        # layer 7 256x32x32
            nn.LeakyReLU(0.01, inplace=True),    # batch norm next
            nn.BatchNorm2d(256),
            nn.MaxPool2d(2),                     # layer 8 256x16x16
            nn.Conv2d(256, 256, 3, 1, 1),        # layer 9 256x16x16
            nn.LeakyReLU(0.01, inplace=True),    # batch norm next
            nn.BatchNorm2d(256),
            nn.Conv2d(256, 1, 3, 1, 1),          # layer 10 1x16x16
        )

    def forward(self, x,y):
        input = torch.cat([x,y], axis=1)
        return self.disc(input)