PyTorch风格迁移算法

本算法主要是基于Pytorch库以及VGG19模型的风格迁移算法，主要依赖的库包括：argparse，torch，torchvision，Pillow。

from __future__ import division
from torchvision import models
from torchvision import transforms
from PIL import Image
import argparse
import torch
import torchvision
import torch.nn as nn
import numpy as np
import mysql.connector

# 设备配置
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

def load_image(image_path, transform=None, max_size=None, shape=None):
    """Load an image and convert it to a torch tensor."""
    image = Image.open(image_path).convert('RGB')
    
    if max_size:
        scale = max_size / max(image.size)
        size = np.array(image.size) * scale
        image = image.resize(size.astype(int), Image.ANTIALIAS)
    
    if shape:
        image = image.resize(shape, Image.LANCZOS)
    
    if transform:
        image = transform(image).unsqueeze(0)
    
    return image.to(device)


class VGGNet(nn.Module):
    def __init__(self):
        """Select conv1_1 ~ conv5_1 activation maps."""
        super(VGGNet, self).__init__()
        self.select = ['0', '5', '10', '19', '28'] 
        self.vgg = models.vgg19(pretrained=True).features
        
    def forward(self, x):
        """Extract multiple convolutional feature maps."""
        features = []
        for name, layer in self.vgg._modules.items():
            x = layer(x)
            if name in self.select:
                features.append(x)
        return features


def main(config):
    
    transform = transforms.Compose([
        transforms.ToTensor(),
        transforms.Normalize(mean=(0.485, 0.456, 0.406), 
                             std=(0.229, 0.224, 0.225))])
    
    # 加载内容与风格图
    # 调整两者为相同大小
    content = load_image(config.content, transform, max_size=config.max_size)
    style = load_image(config.style, transform, shape=[content.size(2), content.size(3)])
    
    # 初始化输出图像（使用内容图）
    target = content.clone().requires_grad_(True)
    
    optimizer = torch.optim.Adam([target], lr=config.lr, betas=[0.5, 0.999])

    vgg = VGGNet().to(device).eval()
    
    for step in range(config.total_step):
        
        # 提取特征向量
        target_features = vgg(target)
        content_features = vgg(content)
        style_features = vgg(style)

        style_loss = 0
        content_loss = 0
        for f1, f2, f3 in zip(target_features, content_features, style_features):
            # 计算内容损失
            content_loss += torch.mean((f1 - f2)**2)

            # 重塑卷积特征映射
            _, c, h, w = f1.size()
            f1 = f1.view(c, h * w)
            f3 = f3.view(c, h * w)

            # 计算gram矩阵
            f1 = torch.mm(f1, f1.t())
            f3 = torch.mm(f3, f3.t())

            # 计算风格损失
            style_loss += torch.mean((f1 - f3)**2) / (c * h * w) 
        
        # 计算总损失并优化
        loss = content_loss + config.style_weight * style_loss 
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        if (step+1) % config.log_step == 0:
            print ('Step [{}/{}], Content Loss: {:.4f}, Style Loss: {:.4f}' 
                   .format(step+1, config.total_step, content_loss.item(), style_loss.item()))

        if (step+1) % config.sample_step == 0:
            # 保存图案
            denorm = transforms.Normalize((-2.12, -2.04, -1.80), (4.37, 4.46, 4.44))
            img = target.clone().squeeze()
            img = denorm(img).clamp_(0, 1)
            torchvision.utils.save_image(img, 'output-{}.png'.format(step+1))
        if(step+1)==config.total_step:
            # 保存图案
            denorm = transforms.Normalize((-2.12, -2.04, -1.80), (4.37, 4.46, 4.44))
            img = target.clone().squeeze()
            img = denorm(img).clamp_(0, 1)
            torchvision.utils.save_image(img, '{}'.format(config.name))


if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument('--content', type=str, default='png/content.png')
    parser.add_argument('--style', type=str, default='png/style.png')
    parser.add_argument('--max_size', type=int, default=400)
    parser.add_argument('--total_step', type=int, default=2000)
    parser.add_argument('--log_step', type=int, default=10)
    parser.add_argument('--sample_step', type=int, default=500)
    parser.add_argument('--style_weight', type=float, default=100)
    parser.add_argument('--lr', type=float, default=0.003)
    parser.add_argument('--user', type=str, default='null')
    parser.add_argument('--name', type=str, default='test')
    config = parser.parse_args()
    print(config)
    torch.cuda.empty_cache()
    main(config)
    mydb = mysql.connector.connect(
    host="",
    user="",
    passwd="",
    database=""
    )
    mycursor = mydb.cursor()
 
    sql = ""
    val = ("", config.name)
 
    mycursor.execute(sql, val)
 
    mydb.commit()
 
    print(mycursor.rowcount, " 条记录被修改")