风格转换简介

风格转换，是把一张图片转化成同内容但包含某风格的新图片。本文将介绍如何让机器学习风格转换，包含两种方法：优化问题求解、转化网络求解。

风格转换

风格转换，就是根据现有的风格照片SS，把当前照片CC转化成带有SS风格同时保留CC内容的照片TT。

本文将叙述两种风格转换的思路：

优化问题

综述

首先，陈述问题：假设已知风格照片SS、当前照片CC，求目标照片TT，要求带有SS的风格并且保留CC的内容。

下面，确定几个损失函数：

最后，便是求解优化问题：

损失函数

优化问题中

是通过预先训练的VGG网络得到。

首先，简单介绍下VGG网络：它是一种固定的网络结构，其结构如下所示，一般采用D或E结构，通常叫VGG-16和VGG-19：

那么，为什么

是通过预先训练的VGG网络得到呢？

训练后的VGG网络，每一层都对特征进行了抽象，越深得到的特征越具象。所以每一层的特征也就代表了图片不同粒度的抽象，可以根据特征的距离判断图片内容的相似程度。VGG的卷积层得到了feature map，假设其大小是C∗H∗W。

训练

构建好损失函数LL后，求解如下优化问题即可：

这里优化问题的求解方法采用L-BFGS（一种伪牛顿法），这样做的目的是得到比gradient descent更快的收敛速度。

例子

本人是詹姆斯的铁杆球迷，对詹姆斯的照片采用不同风格转换后的效果图如下所示。需要说明的是：第二列第一张是未加平滑损失

的效果图，可以看到存在很多噪点，第二列第二张是加入平滑损失

的效果图，照片清晰了很多。

代码

以下代码参考了Siraj Raval on YouTube

代码语言：javascript复制

# Load library
from __future__ import print_function

import time
from PIL import Image
import numpy as np

from keras import backend
from keras.models import Model
from keras.applications.vgg16 import VGG16

from scipy.optimize import fmin_l_bfgs_b
from scipy.misc import imsave

# Load and preprocess the content and style images
height = 512
width = 512

content_image_path = 'images/hugo.jpg'
content_image = Image.open(content_image_path)
content_image = content_image.resize((height, width))
content_image

style_image_path = 'images/styles/wave.jpg'
style_image = Image.open(style_image_path)
style_image = style_image.resize((height, width))
style_image

content_array = np.asarray(content_image, dtype='float32')
content_array = np.expand_dims(content_array, axis=0)
print(content_array.shape)

style_array = np.asarray(style_image, dtype='float32')
style_array = np.expand_dims(style_array, axis=0)
print(style_array.shape)

content_array[:, :, :, 0] -= 103.939
content_array[:, :, :, 1] -= 116.779
content_array[:, :, :, 2] -= 123.68
content_array = content_array[:, :, :, ::-1]

style_array[:, :, :, 0] -= 103.939
style_array[:, :, :, 1] -= 116.779
style_array[:, :, :, 2] -= 123.68
style_array = style_array[:, :, :, ::-1]

content_image = backend.variable(content_array)
style_image = backend.variable(style_array)
combination_image = backend.placeholder((1, height, width, 3))

input_tensor = backend.concatenate([content_image,
                                    style_image,
                                    combination_image], axis=0)

# Reuse a model pre-trained for image classification to define loss functions
model = VGG16(input_tensor=input_tensor, weights='imagenet',
              include_top=False)
layers = dict([(layer.name, layer.output) for layer in model.layers])

content_weight = 0.025
style_weight = 5.0
total_variation_weight = 1.0

# Loss
loss = backend.variable(0.)
# The content loss
def content_loss(content, combination):
    return backend.sum(backend.square(combination - content))

layer_features = layers['block2_conv2']
content_image_features = layer_features[0, :, :, :]
combination_features = layer_features[2, :, :, :]

loss  = content_weight * content_loss(content_image_features,
                                      combination_features)
# The style loss
def gram_matrix(x):
    features = backend.batch_flatten(backend.permute_dimensions(x, (2, 0, 1)))
    gram = backend.dot(features, backend.transpose(features))
    return gram
def style_loss(style, combination):
    S = gram_matrix(style)
    C = gram_matrix(combination)
    channels = 3
    size = height * width
    return backend.sum(backend.square(S - C)) / (4. * (channels ** 2) * (size ** 2))

feature_layers = ['block1_conv2', 'block2_conv2',
                  'block3_conv3', 'block4_conv3',
                  'block5_conv3']
for layer_name in feature_layers:
    layer_features = layers[layer_name]
    style_features = layer_features[1, :, :, :]
    combination_features = layer_features[2, :, :, :]
    sl = style_loss(style_features, combination_features)
    loss  = (style_weight / len(feature_layers)) * sl
# The total variation loss
def total_variation_loss(x):
    a = backend.square(x[:, :height-1, :width-1, :] - x[:, 1:, :width-1, :])
    b = backend.square(x[:, :height-1, :width-1, :] - x[:, :height-1, 1:, :])
    return backend.sum(backend.pow(a   b, 1.25))

loss  = total_variation_weight * total_variation_loss(combination_image)

# Define needed gradients and solve the optimisation problem
grads = backend.gradients(loss, combination_image)
outputs = [loss]
outputs  = grads
f_outputs = backend.function([combination_image], outputs)

def eval_loss_and_grads(x):
    x = x.reshape((1, height, width, 3))
    outs = f_outputs([x])
    loss_value = outs[0]
    grad_values = outs[1].flatten().astype('float64')
    return loss_value, grad_values

class Evaluator(object):

    def __init__(self):
        self.loss_value = None
        self.grads_values = None

    def loss(self, x):
        assert self.loss_value is None
        loss_value, grad_values = eval_loss_and_grads(x)
        self.loss_value = loss_value
        self.grad_values = grad_values
        return self.loss_value

    def grads(self, x):
        assert self.loss_value is not None
        grad_values = np.copy(self.grad_values)
        self.loss_value = None
        self.grad_values = None
        return grad_values

evaluator = Evaluator()

# Train
x = np.random.uniform(0, 255, (1, height, width, 3)) - 128.

iterations = 10

for i in range(iterations):
    print('Start of iteration', i)
    start_time = time.time()
    x, min_val, info = fmin_l_bfgs_b(evaluator.loss, x.flatten(),
                                     fprime=evaluator.grads, maxfun=20)
    print('Current loss value:', min_val)
    end_time = time.time()
    print('Iteration %d completed in %ds' % (i, end_time - start_time))

# Evaluation
x = x.reshape((height, width, 3))
x = x[:, :, ::-1]
x[:, :, 0]  = 103.939
x[:, :, 1]  = 116.779
x[:, :, 2]  = 123.68
x = np.clip(x, 0, 255).astype('uint8')

Image.fromarray(x)