main.py

import argparse
import datetime
import time

from torch.autograd import Variable
from torch.utils.data import DataLoader
from torchvision.utils import save_image

from datasets import *
from u2net import *
from models import *

# os.environ['TF_CPP_MIN_LOG_LEVEL'] = '0'v v v v
os.environ["CUDA_VISIBLE_DEVICES"] = "0"

parser = argparse.ArgumentParser()
parser.add_argument("--epoch", type=int, default=0, help="epoch to start training from")
parser.add_argument("--n_epochs", type=int, default=200, help="number of epochs of training")
parser.add_argument("--dataset_name", type=str, default="New_Data", help="name of the dataset")
parser.add_argument("--model_name", type=str, default="U2netpix", help="name of the model")
parser.add_argument("--batch_size", type=int, default=8, help="size of the batches")
parser.add_argument("--lr", type=float, default=0.0001, help="adam: learning rate")
parser.add_argument("--b1", type=float, default=0.5, help="adam: decay of first order momentum of gradient")
parser.add_argument("--b2", type=float, default=0.999, help="adam: decay of first order momentum of gradient")
parser.add_argument("--decay_epoch", type=int, default=100, help="epoch from which to start lr decay")
parser.add_argument("--n_cpu", type=int, default=8, help="number of cpu threads to use during batch generation")
parser.add_argument("--img_height", type=int, default=512, help="size of image height")
parser.add_argument("--img_width", type=int, default=512, help="size of image width")
parser.add_argument("--channels", type=int, default=3, help="number of image channels")
parser.add_argument("--sample_interval", type=int, default=100, help="interval between sampling of images from generators")
parser.add_argument("--checkpoint_interval", type=int, default=50, help="interval between model checkpoints")
opt = parser.parse_args()
print(opt)

os.makedirs("images/%s" % opt.model_name, exist_ok=True)
os.makedirs("saved_models/%s" % opt.model_name, exist_ok=True)

cuda = True if torch.cuda.is_available() else False

# Loss functions
criterion_GAN = torch.nn.MSELoss()
criterion_pixelwise = torch.nn.L1Loss()


def mutil_mutil_criterion_pixelwise(d0, GT):
    loss0 = criterion_pixelwise(d0, GT)


    loss = loss0
    # print("\nl0: %3f, l1: %3f, l2: %3f, l3: %3f, l4: %3f, l5: %3f, l6: %3f\n" % (
    #     loss0.item(), loss1.item(), loss2.item(), loss3.item(), loss4.item(), loss5.item(), loss6.item()))

    return loss0, loss


class ValidateDataset(Dataset):

    def __init__(self, root, transforms_=None, mode="train"):
        self.transform = transforms.Compose(transforms_)

        self.files = root

    def __getitem__(self, index):
        img = Image.open(self.files[index % len(self.files)])
        w, h = img.size
        img_A = img.crop((0, 0, w / 2, h))
        img_B = img.crop((w / 2, 0, w, h))

        img_A = self.transform(img_A)
        img_B = self.transform(img_B)

        return {"A": img_A, "B": img_B}

    def __len__(self):
        return len(self.files)


# Loss weight of L1 pixel-wise loss between translated image and real image
lambda_pixel = 100

# Calculate output of image discriminator (PatchGAN)
patch = (1, opt.img_height // 2 ** 4, opt.img_width // 2 ** 4)

# Initialize generator and discriminator
generator = GeneratorUNet(3, 3)
discriminator = Discriminator()

if cuda:
    generator = generator.cuda()
    discriminator = discriminator.cuda()
    criterion_GAN.cuda()
    criterion_pixelwise.cuda()

if opt.epoch != 0:
    # Load pretrained models
    generator.load_state_dict(torch.load("saved_models/%s/generator_%d.pth" % (opt.model_name, opt.epoch)))
    discriminator.load_state_dict(torch.load("saved_models/%s/discriminator_%d.pth" % (opt.model_name, opt.epoch)))
else:
    # Initialize weights
    generator.apply(weights_init_normal)
    discriminator.apply(weights_init_normal)

# Optimizers
optimizer_G = torch.optim.Adam(generator.parameters(), lr=opt.lr, betas=(opt.b1, opt.b2))
optimizer_D = torch.optim.Adam(discriminator.parameters(), lr=opt.lr, betas=(opt.b1, opt.b2))

# Configure dataloaders
transforms_ = [
    transforms.Resize((opt.img_height, opt.img_width), Image.BICUBIC),
    transforms.ToTensor(),
    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
]

dataloader = DataLoader(
    ImageDataset("./data/%s" % opt.dataset_name, transforms_=transforms_),
    batch_size=opt.batch_size,
    shuffle=True,
    # num_workers=opt.n_cpu,
)

val_dataloader = DataLoader(
    ImageDataset("./data/%s" % opt.dataset_name, transforms_=transforms_, mode="val"),
    batch_size=3,
    shuffle=True,
)

# Tensor type
Tensor = torch.cuda.FloatTensor if cuda else torch.FloatTensor


def sample_images(batches_done):
    """Saves a generated sample from the validation set"""
    imgs = next(iter(val_dataloader))
    real_A = Variable(imgs["B"].type(Tensor))
    real_B = Variable(imgs["A"].type(Tensor))
    fake_B = generator(real_A)
    img_sample = torch.cat((real_A.data, fake_B.data, real_B.data), -1)
    save_image(img_sample, "images/%s/%s.png" % (opt.model_name, batches_done), nrow=1, padding=0, normalize=True)

# ----------
#  Training
# ----------

prev_time = time.time()

for epoch in range(opt.epoch, opt.n_epochs):
    for i, batch in enumerate(dataloader):

        # Model inputs
        real_A = Variable(batch["B"].type(Tensor))
        real_B = Variable(batch["A"].type(Tensor))

        # Adversarial ground truths
        valid = Variable(Tensor(np.ones((real_A.size(0), *patch))), requires_grad=False)
        fake = Variable(Tensor(np.zeros((real_A.size(0), *patch))), requires_grad=False)

        # ------------------
        #  Train Generators
        # ------------------

        optimizer_G.zero_grad()

        # GAN loss
        fake_B0 = generator(real_A)
        pred_fake = discriminator(fake_B0, real_A)
        loss_GAN = criterion_GAN(pred_fake, valid)
        # Pixel-wise loss
        loss_pixel0, loss_pixel_all = mutil_mutil_criterion_pixelwise(fake_B0, real_B)

        # Total loss
        loss_G = loss_GAN + lambda_pixel * (loss_pixel_all / 7)
        # loss_G = loss_GAN + loss_pixel_all
        # with torch.no_grad():
        loss_G.backward()

        optimizer_G.step()

        # ---------------------
        #  Train Discriminator
        # ---------------------

        optimizer_D.zero_grad()

        # Real loss
        pred_real = discriminator(real_B, real_A)
        loss_real = criterion_GAN(pred_real, valid)

        # Fake loss
        pred_fake = discriminator(fake_B0.detach(), real_A)
        loss_fake = criterion_GAN(pred_fake, fake)

        # Total loss
        loss_D = 0.5 * (loss_real + loss_fake)

        loss_D.backward()
        optimizer_D.step()

        # Determine approximate time left
        batches_done = epoch * len(dataloader) + i
        batches_left = opt.n_epochs * len(dataloader) - batches_done
        time_left = datetime.timedelta(seconds=batches_left * (time.time() - prev_time))
        prev_time = time.time()

        # Print log
        # sys.stdout.write(
        print(
            "\r[Epoch %d/%d] [Batch %d/%d] [D loss: %f] [G loss: %f, pixel: %f, adv: %f] ETA: %s"
            % (
                epoch,
                opt.n_epochs,
                i,
                len(dataloader),
                loss_D.item(),
                loss_G.item(),
                loss_pixel0.item(),
                loss_GAN.item(),
                time_left,
            )
        )

        # If at sample interval save image

        if batches_done % opt.sample_interval == 0:
            sample_images(epoch)

    if epoch > 399 and epoch % opt.checkpoint_interval == 0:
        # Save model checkpoints
        torch.save(generator.state_dict(), "saved_models/%s/generator_%d.pth" % (opt.model_name, epoch))
        torch.save(discriminator.state_dict(), "saved_models/%s/discriminator_%d.pth" % (opt.model_name, epoch))