model.py

import torch
import torch.nn as nn
import torch.nn.functional as F


class Discriminator(nn.Module):
    def __init__(self, dropout_prob=0.3):
        super(Discriminator, self).__init__()

        self.main = nn.Sequential(
            nn.Conv2d(3, 64, 4, stride=2, padding=1),
            nn.LeakyReLU(0.2, inplace=True),
            nn.Dropout2d(p=dropout_prob),

            nn.Conv2d(64, 128, 4, stride=2, padding=1),
            nn.InstanceNorm2d(128),
            nn.LeakyReLU(0.2, inplace=True),
            nn.Dropout2d(p=dropout_prob),

            nn.Conv2d(128, 256, 4, stride=2, padding=1),
            nn.InstanceNorm2d(256),
            nn.LeakyReLU(0.2, inplace=True),
            nn.Dropout2d(p=dropout_prob),

            nn.Conv2d(256, 512, 4, padding=1),
            nn.InstanceNorm2d(512),
            nn.LeakyReLU(0.2, inplace=True),
            nn.Dropout2d(p=dropout_prob),

            nn.Conv2d(512, 1, 4, padding=1),
        )

    def forward(self, x):
        x = self.main(x)
        x = F.avg_pool2d(x, x.size()[2:])
        x = torch.flatten(x, 1)
        x = torch.sigmoid(x)
        return x


class Generator(nn.Module):
    def __init__(self):
        super(Generator, self).__init__()
        self.main = nn.Sequential(
            # Initial convolution block
            nn.ReflectionPad2d(3),
            nn.Conv2d(3, 64, 7),
            nn.InstanceNorm2d(64),
            nn.ReLU(inplace=True),

            # Downsampling
            nn.Conv2d(64, 128, 3, stride=2, padding=1),
            nn.InstanceNorm2d(128),
            nn.ReLU(inplace=True),
            nn.Conv2d(128, 256, 3, stride=2, padding=1),
            nn.InstanceNorm2d(256),
            nn.ReLU(inplace=True),

            # Residual blocks
            ResidualBlock(256),
            ResidualBlock(256),
            ResidualBlock(256),
            ResidualBlock(256),
            ResidualBlock(256),
            ResidualBlock(256),
            ResidualBlock(256),
            ResidualBlock(256),
            ResidualBlock(256),

            # Upsampling
            nn.ConvTranspose2d(256, 128, 3, stride=2, padding=1, output_padding=1),
            nn.InstanceNorm2d(128),
            nn.ReLU(inplace=True),
            nn.ConvTranspose2d(128, 64, 3, stride=2, padding=1, output_padding=1),
            nn.InstanceNorm2d(64),
            nn.ReLU(inplace=True),

            # Output layer
            nn.ReflectionPad2d(3),
            nn.Conv2d(64, 3, 7),
            nn.Tanh()
        )

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


class ResidualBlock(nn.Module):
    def __init__(self, in_channels):
        super(ResidualBlock, self).__init__()

        self.res = nn.Sequential(nn.ReflectionPad2d(1),
                                 nn.Conv2d(in_channels, in_channels, 3),
                                 nn.InstanceNorm2d(in_channels),
                                 nn.ReLU(inplace=True),
                                 nn.ReflectionPad2d(1),
                                 nn.Conv2d(in_channels, in_channels, 3),
                                 nn.InstanceNorm2d(in_channels))

    def forward(self, x):
        return x + self.res(x)


class CycleGAN(nn.Module):
    def __init__(self, descriminator, generator):
        super(CycleGAN, self).__init__()

        self.netG_A2B = generator()
        self.netG_B2A = generator()
        self.netD_A   = descriminator()
        self.netD_B   = descriminator()