【源码解读】cycleGAN（一）：网络

源码地址：https://github.com/aitorzip/PyTorch-CycleGAN

如图所示，cycleGAN的网络结构包括两个生成器G(X->Y)和F(Y->X)，两个判别器Dx和Dy

生成器部分：网络整体上经过一个降采样然后上采样的过程，中间是一系列残差块,数目由实际情况确定，根据论文中所说，当输入分辨率为128x128，采用6个残差块，当输入分辨率为256x256甚至更高时，采用9个残差块，其源代码如下,

class Generator(nn.Module):
    def __init__(self, input_nc, output_nc, n_residual_blocks=9):
        super(Generator, self).__init__()

        # Initial convolution block
        model = [   nn.ReflectionPad2d(3),
                    nn.Conv2d(input_nc, 64, 7),
                    nn.InstanceNorm2d(64),
                    nn.ReLU(inplace=True) ]

        # Downsampling
        in_features = 64
        out_features = in_features*2
        for _ in range(2):
            model += [  nn.Conv2d(in_features, out_features, 3, stride=2, padding=1),
                        nn.InstanceNorm2d(out_features),
                        nn.ReLU(inplace=True) ]
            in_features = out_features
            out_features = in_features*2

        # Residual blocks
        for _ in range(n_residual_blocks):
            model += [ResidualBlock(in_features)]

        # Upsampling
        out_features = in_features//2
        for _ in range(2):
            model += [  nn.ConvTranspose2d(in_features, out_features, 3, stride=2, padding=1, output_padding=1),
                        nn.InstanceNorm2d(out_features),
                        nn.ReLU(inplace=True) ]
            in_features = out_features
            out_features = in_features//2

        # Output layer
        model += [  nn.ReflectionPad2d(3),
                    nn.Conv2d(64, output_nc, 7),
                    nn.Tanh() ]

        self.model = nn.Sequential(*model)

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

其中，值得注意的网络层是nn.ReflectionPad2d和nn.InstanceNorm2d，前者搭配7x7卷积，先在特征图周围以反射的方式补长度，使得卷积后特征图尺寸不变，示例如下，输出结果就是以特征图边界为反射边，向外补充

nn.InstanceNorm2d是相比于batchNorm更加适合图像生成，风格迁移的归一化方法，相比于batchNorm跨样本，单通道统计，InstanceNorm采用单样本，单通道统计，括号中的参数代表通道数。

判别器部分：结构比生成器更加简单，经过5层卷积，通道数缩减为1，最后池化平均，尺寸也缩减为1x1，最最后reshape一下，变为（batchsize,1)

class Discriminator(nn.Module):
    def __init__(self, input_nc):
        super(Discriminator, self).__init__()

        # A bunch of convolutions one after another
        model = [   nn.Conv2d(input_nc, 64, 4, stride=2, padding=1),
                    nn.LeakyReLU(0.2, inplace=True) ]

        model += [  nn.Conv2d(64, 128, 4, stride=2, padding=1),
                    nn.InstanceNorm2d(128),
                    nn.LeakyReLU(0.2, inplace=True) ]

        model += [  nn.Conv2d(128, 256, 4, stride=2, padding=1),
                    nn.InstanceNorm2d(256),
                    nn.LeakyReLU(0.2, inplace=True) ]

        model += [  nn.Conv2d(256, 512, 4, padding=1),
                    nn.InstanceNorm2d(512),
                    nn.LeakyReLU(0.2, inplace=True) ]

        # FCN classification layer
        model += [nn.Conv2d(512, 1, 4, padding=1)]

        self.model = nn.Sequential(*model)

    def forward(self, x):
        x =  self.model(x)
        # Average pooling and flatten
        return F.avg_pool2d(x, x.size()[2:]).view(x.size()[0])