Neural style transfer (NST) test – success

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

from PIL import Image
import matplotlib.pyplot as plt

import torchvision.transforms as transforms
import torchvision.models as models

import copy

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

# Define the fixed size
target_size = (512, 512)

loader = transforms.Compose([
    transforms.Resize(target_size),
    transforms.ToTensor()])

def image_loader(image_name):
    image = Image.open(image_name)
    image = loader(image).unsqueeze(0)
    return image.to(device, torch.float)

style_img = image_loader("style_fanGao.jpg")
content_img = image_loader("Zengbo.jpg")

# Resize the images to the same size
style_img = F.interpolate(style_img, size=target_size, mode='bilinear', align_corners=False)
content_img = F.interpolate(content_img, size=target_size, mode='bilinear', align_corners=False)

assert style_img.size() == content_img.size(), "Style and content images must be of the same size."

unloader = transforms.ToPILImage()

plt.ion()

def imshow(tensor, title=None):
    image = tensor.cpu().clone()
    image = image.squeeze(0)
    image = unloader(image)
    plt.imshow(image)
    if title is not None:
        plt.title(title)
    plt.pause(0.001)

def gram_matrix(input):
    a, b, c, d = input.size()  # a=batch size(=1)
    features = input.view(a * b, c * d)  # resise F_XL into \hat F_XL

    G = torch.mm(features, features.t())  # compute the gram product

    # normalization factor
    return G.div(a * b * c * d)

class ContentLoss(nn.Module):
    def __init__(self, target):
        super(ContentLoss, self).__init__()
        self.target = target.detach()

    def forward(self, input):
        self.loss = F.mse_loss(input, self.target)
        return input

class StyleLoss(nn.Module):
    def __init__(self, target_feature):
        super(StyleLoss, self).__init__()
        self.target = gram_matrix(target_feature).detach()

    def forward(self, input):
        G = gram_matrix(input)
        self.loss = F.mse_loss(G, self.target)
        return input

class Normalization(nn.Module):
    def __init__(self, mean, std):
        super(Normalization, self).__init__()
        self.mean = mean.view(-1, 1, 1)
        self.std = std.view(-1, 1, 1)

    def forward(self, img):
        return (img - self.mean) / self.std

cnn = models.vgg19(pretrained=True).features.to(device).eval()

cnn_normalization_mean = torch.tensor([0.485, 0.456, 0.406]).to(device)
cnn_normalization_std = torch.tensor([0.229, 0.224, 0.225]).to(device)

content_layers_default = ['conv_4']
style_layers_default = ['conv_1', 'conv_2', 'conv_3', 'conv_4', 'conv_5']

def get_style_model_and_losses(cnn, normalization_mean, normalization_std,
                               style_img, content_img,
                               content_layers=content_layers_default,
                               style_layers=style_layers_default):
    normalization = Normalization(normalization_mean, normalization_std).to(device)

    content_losses = []
    style_losses = []

    model = nn.Sequential(normalization)

    i = 0
    for layer in cnn.children():
        if isinstance(layer, nn.Conv2d):
            i += 1
            name = 'conv_{}'.format(i)
        elif isinstance(layer, nn.ReLU):
            name = 'relu_{}'.format(i)
            layer = nn.ReLU(inplace=False)
        elif isinstance(layer, nn.MaxPool2d):
            name = 'pool_{}'.format(i)
        elif isinstance(layer, nn.BatchNorm2d):
            name = 'bn_{}'.format(i)
        else:
            raise RuntimeError('Unrecognized layer: {}'.format(layer.__class__.__name__))

        model.add_module(name, layer)

        if name in content_layers:
            target = model(content_img).detach()
            content_loss = ContentLoss(target)
            model.add_module("content_loss_{}".format(i), content_loss)
            content_losses.append(content_loss)

        if name in style_layers:
            target_feature = model(style_img).detach()
            style_loss = StyleLoss(target_feature)
            model.add_module("style_loss_{}".format(i), style_loss)
            style_losses.append(style_loss)

    for i in range(len(model) - 1, -1, -1):
        if isinstance(model[i], ContentLoss) or isinstance(model[i], StyleLoss):
            break

    model = model[:(i + 1)]

    return model, style_losses, content_losses

def get_input_optimizer(input_img):
    optimizer = optim.LBFGS([input_img])
    return optimizer

def run_style_transfer(cnn, normalization_mean, normalization_std,
                       content_img, style_img, input_img, num_steps=300,
                       style_weight=1000000, content_weight=1):
    model, style_losses, content_losses = get_style_model_and_losses(cnn,
                                                                      normalization_mean, normalization_std,
                                                                      style_img, content_img)

    input_img.requires_grad_(True)
    model.requires_grad_(False)

    optimizer = get_input_optimizer(input_img)

    run = [0]
    while run[0] <= num_steps:

        def closure():
            with torch.no_grad():
                input_img.clamp_(0, 1)

            optimizer.zero_grad()
            model(input_img)
            style_score = 0
            content_score = 0

            for sl in style_losses:
                style_score += sl.loss
            for cl in content_losses:
                content_score += cl.loss

            style_score *= style_weight
            content_score *= content_weight

            loss = style_score + content_score
            loss.backward()

            run[0] += 1
            if run[0] % 50 == 0:
                print("run {}:".format(run))
                print('Style Loss : {:4f} Content Loss: {:4f}'.format(
                    style_score.item(), content_score.item()))
                print()

            return style_score + content_score

        optimizer.step(closure)

    with torch.no_grad():
        input_img.clamp_(0, 1)

    return input_img

def save_output_image(image, output_path):
    try:
        image.save(output_path)
        print(f"Output image saved successfully at {output_path}")
    except Exception as e:
        print(f"An error occurred while saving the output image: {str(e)}")

def convert_to_pil_image(tensor):
    image = tensor.squeeze(0).cpu().clone().detach().numpy().transpose(1, 2, 0)
    image = image.clip(0, 1)
    image = (image * 255).astype('uint8')
    return Image.fromarray(image)

num_steps = 500
style_weight = 1000000000
content_weight = 1

input_img = content_img.clone()

output = run_style_transfer(cnn, cnn_normalization_mean, cnn_normalization_std,
                            content_img, style_img, input_img, num_steps=num_steps,
                            style_weight=style_weight, content_weight=content_weight)

output_image = convert_to_pil_image(output)

output_path = "output_image.jpg"
save_output_image(output_image, output_path)

# Display content, style, and output images in a single row
fig, axes = plt.subplots(1, 3, figsize=(15, 5))
axes[0].imshow(content_img.squeeze(0).permute(1, 2, 0).cpu().numpy())
axes[0].set_title('Content Image')
axes[0].axis('off')
axes[1].imshow(style_img.squeeze(0).permute(1, 2, 0).cpu().numpy())
axes[1].set_title('Style Image')
axes[1].axis('off')
axes[2].imshow(output_image)
axes[2].set_title('Output Image')
axes[2].axis('off')
plt.tight_layout()
plt.show()

Leave a Reply