cyclegan.py

from ..torch_core import *
from ..layers import *
from ..callback import *
from ..basic_train import Learner, LearnerCallback

__all__ = ['CycleGAN', 'CycleGanLoss', 'AdaptiveLoss', 'CycleGANTrainer']

def convT_norm_relu(ch_in:int, ch_out:int, norm_layer:nn.Module, ks:int=3, stride:int=2, bias:bool=True):
    return [nn.ConvTranspose2d(ch_in, ch_out, kernel_size=ks, stride=stride, padding=1, output_padding=1, bias=bias),
            norm_layer(ch_out), nn.ReLU(True)]

def pad_conv_norm_relu(ch_in:int, ch_out:int, pad_mode:str, norm_layer:nn.Module, ks:int=3, bias:bool=True,
                       pad=1, stride:int=1, activ:bool=True, init:Callable=nn.init.kaiming_normal_)->List[nn.Module]:
    layers = []
    if pad_mode == 'reflection': layers.append(nn.ReflectionPad2d(pad))
    elif pad_mode == 'border':   layers.append(nn.ReplicationPad2d(pad))
    p = pad if pad_mode == 'zeros' else 0
    conv = nn.Conv2d(ch_in, ch_out, kernel_size=ks, padding=p, stride=stride, bias=bias)
    if init:
        init(conv.weight)
        if hasattr(conv, 'bias') and hasattr(conv.bias, 'data'): conv.bias.data.fill_(0.)
    layers += [conv, norm_layer(ch_out)]
    if activ: layers.append(nn.ReLU(inplace=True))
    return layers

class ResnetBlock(Module):
    def __init__(self, dim:int, pad_mode:str='reflection', norm_layer:nn.Module=None, dropout:float=0., bias:bool=True):
        assert pad_mode in ['zeros', 'reflection', 'border'], f'padding {pad_mode} not implemented.'
        norm_layer = ifnone(norm_layer, nn.InstanceNorm2d)
        layers = pad_conv_norm_relu(dim, dim, pad_mode, norm_layer, bias=bias)
        if dropout != 0: layers.append(nn.Dropout(dropout))
        layers += pad_conv_norm_relu(dim, dim, pad_mode, norm_layer, bias=bias, activ=False)
        self.conv_block = nn.Sequential(*layers)

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

def resnet_generator(ch_in:int, ch_out:int, n_ftrs:int=64, norm_layer:nn.Module=None,
                     dropout:float=0., n_blocks:int=6, pad_mode:str='reflection')->nn.Module:
    norm_layer = ifnone(norm_layer, nn.InstanceNorm2d)
    bias = (norm_layer == nn.InstanceNorm2d)
    layers = pad_conv_norm_relu(ch_in, n_ftrs, 'reflection', norm_layer, pad=3, ks=7, bias=bias)
    for i in range(2):
        layers += pad_conv_norm_relu(n_ftrs, n_ftrs *2, 'zeros', norm_layer, stride=2, bias=bias)
        n_ftrs *= 2
    layers += [ResnetBlock(n_ftrs, pad_mode, norm_layer, dropout, bias) for _ in range(n_blocks)]
    for i in range(2):
        layers += convT_norm_relu(n_ftrs, n_ftrs//2, norm_layer, bias=bias)
        n_ftrs //= 2
    layers += [nn.ReflectionPad2d(3), nn.Conv2d(n_ftrs, ch_out, kernel_size=7, padding=0), nn.Tanh()]
    return nn.Sequential(*layers)

def conv_norm_lr(ch_in:int, ch_out:int, norm_layer:nn.Module=None, ks:int=3, bias:bool=True, pad:int=1, stride:int=1,
                 activ:bool=True, slope:float=0.2, init:Callable=nn.init.kaiming_normal_)->List[nn.Module]:
    conv = nn.Conv2d(ch_in, ch_out, kernel_size=ks, padding=pad, stride=stride, bias=bias)
    if init:
        init(conv.weight)
        if hasattr(conv, 'bias') and hasattr(conv.bias, 'data'): conv.bias.data.fill_(0.)
    layers = [conv]
    if norm_layer is not None: layers.append(norm_layer(ch_out))
    if activ: layers.append(nn.LeakyReLU(slope, inplace=True))
    return layers

def critic(ch_in:int, n_ftrs:int=64, n_layers:int=3, norm_layer:nn.Module=None, sigmoid:bool=False)->nn.Module:
    norm_layer = ifnone(norm_layer, nn.InstanceNorm2d)
    bias = (norm_layer == nn.InstanceNorm2d)
    layers = conv_norm_lr(ch_in, n_ftrs, ks=4, stride=2, pad=1)
    for i in range(n_layers-1):
        new_ftrs = 2*n_ftrs if i <= 3 else n_ftrs
        layers += conv_norm_lr(n_ftrs, new_ftrs, norm_layer, ks=4, stride=2, pad=1, bias=bias)
        n_ftrs = new_ftrs
    new_ftrs = 2*n_ftrs if n_layers <=3 else n_ftrs
    layers += conv_norm_lr(n_ftrs, new_ftrs, norm_layer, ks=4, stride=1, pad=1, bias=bias)
    layers.append(nn.Conv2d(new_ftrs, 1, kernel_size=4, stride=1, padding=1))
    if sigmoid: layers.append(nn.Sigmoid())
    return nn.Sequential(*layers)

class CycleGAN(Module):

    def __init__(self, ch_in:int, ch_out:int, n_features:int=64, disc_layers:int=3, gen_blocks:int=6, lsgan:bool=True,
                 drop:float=0., norm_layer:nn.Module=None):
        self.D_A = critic(ch_in, n_features, disc_layers, norm_layer, sigmoid=not lsgan)
        self.D_B = critic(ch_in, n_features, disc_layers, norm_layer, sigmoid=not lsgan)
        self.G_A = resnet_generator(ch_in, ch_out, n_features, norm_layer, drop, gen_blocks)
        self.G_B = resnet_generator(ch_in, ch_out, n_features, norm_layer, drop, gen_blocks)
        #G_A: takes real input B and generates fake input A
        #G_B: takes real input A and generates fake input B
        #D_A: trained to make the difference between real input A and fake input A
        #D_B: trained to make the difference between real input B and fake input B

    def forward(self, real_A, real_B):
        fake_A, fake_B = self.G_A(real_B), self.G_B(real_A)
        if not self.training: return torch.cat([fake_A[:,None],fake_B[:,None]], 1)
        idt_A, idt_B = self.G_A(real_A), self.G_B(real_B)
        return [fake_A, fake_B, idt_A, idt_B]

class AdaptiveLoss(Module):
    def __init__(self, crit): self.crit = crit

    def forward(self, output, target:bool):
        targ = output.new_ones(*output.size()) if target else output.new_zeros(*output.size())
        return self.crit(output, targ)

class CycleGanLoss(Module):
    def __init__(self, cgan:nn.Module, lambda_A:float=10., lambda_B:float=10, lambda_idt:float=0.5, lsgan:bool=True):
        self.cgan,self.l_A,self.l_B,self.l_idt = cgan,lambda_A,lambda_B,lambda_idt
        #self.crit = F.mse_loss if lsgan else F.binary_cross_entropy
        self.crit = AdaptiveLoss(F.mse_loss if lsgan else F.binary_cross_entropy)

    def set_input(self, input):
        self.real_A,self.real_B = input

    def forward(self, output, target):
        fake_A, fake_B, idt_A, idt_B = output
        #Generators should return identity on the datasets they try to convert to
        idt_loss = self.l_idt * (self.l_B * F.l1_loss(idt_A, self.real_B) + self.l_A * F.l1_loss(idt_B, self.real_A))
        #Generators are trained to trick the critics so the following should be ones
        gen_loss = self.crit(self.cgan.D_A(fake_A), True) + self.crit(self.cgan.D_B(fake_B), True)
        #Cycle loss
        cycle_loss = self.l_A * F.l1_loss(self.cgan.G_A(fake_B), self.real_A)
        cycle_loss += self.l_B * F.l1_loss(self.cgan.G_B(fake_A), self.real_B)
        self.metrics = [idt_loss, gen_loss, cycle_loss]
        return idt_loss + gen_loss + cycle_loss

class CycleGANTrainer(LearnerCallback):
    "`LearnerCallback` that handles cycleGAN Training."
    _order=-20
    def _set_trainable(self, D_A=False, D_B=False):
        gen = (not D_A) and (not D_B)
        requires_grad(self.learn.model.G_A, gen)
        requires_grad(self.learn.model.G_B, gen)
        requires_grad(self.learn.model.D_A, D_A)
        requires_grad(self.learn.model.D_B, D_B)
        if not gen:
            self.opt_D_A.lr, self.opt_D_A.mom = self.learn.opt.lr, self.learn.opt.mom
            self.opt_D_A.wd, self.opt_D_A.beta = self.learn.opt.wd, self.learn.opt.beta
            self.opt_D_B.lr, self.opt_D_B.mom = self.learn.opt.lr, self.learn.opt.mom
            self.opt_D_B.wd, self.opt_D_B.beta = self.learn.opt.wd, self.learn.opt.beta

    def on_train_begin(self, **kwargs):
        "Create the various optimizers."
        self.G_A,self.G_B = self.learn.model.G_A,self.learn.model.G_B
        self.D_A,self.D_B = self.learn.model.D_A,self.learn.model.D_B
        self.crit = self.learn.loss_func.crit
        self.opt_G = self.learn.opt.new([nn.Sequential(*flatten_model(self.G_A), *flatten_model(self.G_B))])
        self.opt_D_A = self.learn.opt.new([nn.Sequential(*flatten_model(self.D_A))])
        self.opt_D_B = self.learn.opt.new([nn.Sequential(*flatten_model(self.D_B))])
        self.learn.opt.opt = self.opt_G.opt
        self._set_trainable()
        self.names = ['idt_loss', 'gen_loss', 'cyc_loss', 'da_loss', 'db_loss']
        self.learn.recorder.no_val=True
        self.learn.recorder.add_metric_names(self.names)
        self.smootheners = {n:SmoothenValue(0.98) for n in self.names}

    def on_batch_begin(self, last_input, **kwargs):
        "Register the `last_input` in the loss function."
        self.learn.loss_func.set_input(last_input)

    def on_batch_end(self, last_input, last_output, **kwargs):
        "Steps through the generators then each of the critics."
        self.G_A.zero_grad(); self.G_B.zero_grad()
        fake_A, fake_B = last_output[0].detach(), last_output[1].detach()
        real_A, real_B = last_input
        self._set_trainable(D_A=True)
        self.D_A.zero_grad()
        loss_D_A = 0.5 * (self.crit(self.D_A(real_A), True) + self.crit(self.D_A(fake_A), False))
        loss_D_A.backward()
        self.opt_D_A.step()
        self._set_trainable(D_B=True)
        self.D_B.zero_grad()
        loss_D_B = 0.5 * (self.crit(self.D_B(real_B), True) + self.crit(self.D_B(fake_B), False))
        loss_D_B.backward()
        self.opt_D_B.step()
        self._set_trainable()
        metrics = self.learn.loss_func.metrics + [loss_D_A, loss_D_B]
        for n,m in zip(self.names,metrics): self.smootheners[n].add_value(m)

    def on_epoch_end(self, last_metrics, **kwargs):
        "Put the various losses in the recorder."
        return add_metrics(last_metrics, [s.smooth for k,s in self.smootheners.items()])