在上一篇博客里,我们介绍了利用卷积神经网络实现图像风格迁移的算法原理, 这篇文章,我们主要介绍基于TensorFlow的程序实现,为了实现以下程序,你需要安装 TensorFlow, Numpy, Scipy, 以及下载 VGG-19 model。
参考来源:
import osimport sysimport numpy as npimport scipy.ioimport scipy.miscimport tensorflow as tf# Output folder for the images.OUTPUT_DIR = 'output/'# Style image to use.STYLE_IMAGE = '/images/ocean.jpg'# Content image to use.CONTENT_IMAGE = '/images/Taipei101.jpg'# Image dimensions constants.IMAGE_WIDTH = 800IMAGE_HEIGHT = 600COLOR_CHANNELS = 3################################################################################ Algorithm constants################################################################################ 设置随机噪声图像与内容图像的比率NOISE_RATIO = 0.6# 设置迭代次数ITERATIONS = 1000# 设置内容图像与风格图像的权重alpha = 1beta = 500# 加载VGG-19 MODEL及设定均值VGG_Model = 'Downloads/imagenet-vgg-verydeep-19.mat'MEAN_VALUES = np.array([123.68, 116.779, 103.939]).reshape((1, 1, 1, 3))# 设置需要用到的卷积层CONTENT_LAYERS = [('conv4_2', 1.)]STYLE_LAYERS = [('conv1_1', 0.2), ('conv2_1', 0.2), ('conv3_1', 0.2), ('conv4_1', 0.2), ('conv5_1', 0.2)]# 生成随机噪声图,与content图以一定比率融合def generate_noise_image(content_image, noise_ratio = NOISE_RATIO): """ Returns a noise image intermixed with the content image at a certain ratio. """ noise_image = np.random.uniform( -20, 20, (1, IMAGE_HEIGHT, IMAGE_WIDTH, COLOR_CHANNELS)).astype('float32') # White noise image from the content representation. Take a weighted average # of the values img = noise_image * noise_ratio + content_image * (1 - noise_ratio) return imgdef load_image(path): image = scipy.misc.imread(path) # Resize the image for convnet input, there is no change but just # add an extra dimension. image = np.reshape(image, ((1,) + image.shape)) # Input to the VGG net expects the mean to be subtracted. image = image - MEAN_VALUES return imagedef save_image(path, image): # Output should add back the mean. image = image + MEAN_VALUES # Get rid of the first useless dimension, what remains is the image. image = image[0] image = np.clip(image, 0, 255).astype('uint8') scipy.misc.imsave(path, image)def build_net(ntype, nin, nwb=None): if ntype == 'conv': return tf.nn.relu(tf.nn.conv2d(nin, nwb[0], strides=[1, 1, 1, 1], padding='SAME') + nwb[1]) elif ntype == 'pool': return tf.nn.avg_pool(nin, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')def get_weight_bias(vgg_layers, i): weights = vgg_layers[i][0][0][2][0][0] weights = tf.constant(weights) bias = vgg_layers[i][0][0][2][0][1] bias = tf.constant(np.reshape(bias, (bias.size))) return weights, biasdef build_vgg19(path): net = {} vgg_rawnet = scipy.io.loadmat(path) vgg_layers = vgg_rawnet['layers'][0] net['input'] = tf.Variable(np.zeros((1, IMAGE_HEIGHT, IMAGE_WIDTH, 3)).astype('float32')) net['conv1_1'] = build_net('conv', net['input'], get_weight_bias(vgg_layers, 0)) net['conv1_2'] = build_net('conv', net['conv1_1'], get_weight_bias(vgg_layers, 2)) net['pool1'] = build_net('pool', net['conv1_2']) net['conv2_1'] = build_net('conv', net['pool1'], get_weight_bias(vgg_layers, 5)) net['conv2_2'] = build_net('conv', net['conv2_1'], get_weight_bias(vgg_layers, 7)) net['pool2'] = build_net('pool', net['conv2_2']) net['conv3_1'] = build_net('conv', net['pool2'], get_weight_bias(vgg_layers, 10)) net['conv3_2'] = build_net('conv', net['conv3_1'], get_weight_bias(vgg_layers, 12)) net['conv3_3'] = build_net('conv', net['conv3_2'], get_weight_bias(vgg_layers, 14)) net['conv3_4'] = build_net('conv', net['conv3_3'], get_weight_bias(vgg_layers, 16)) net['pool3'] = build_net('pool', net['conv3_4']) net['conv4_1'] = build_net('conv', net['pool3'], get_weight_bias(vgg_layers, 19)) net['conv4_2'] = build_net('conv', net['conv4_1'], get_weight_bias(vgg_layers, 21)) net['conv4_3'] = build_net('conv', net['conv4_2'], get_weight_bias(vgg_layers, 23)) net['conv4_4'] = build_net('conv', net['conv4_3'], get_weight_bias(vgg_layers, 25)) net['pool4'] = build_net('pool', net['conv4_4']) net['conv5_1'] = build_net('conv', net['pool4'], get_weight_bias(vgg_layers, 28)) net['conv5_2'] = build_net('conv', net['conv5_1'], get_weight_bias(vgg_layers, 30)) net['conv5_3'] = build_net('conv', net['conv5_2'], get_weight_bias(vgg_layers, 32)) net['conv5_4'] = build_net('conv', net['conv5_3'], get_weight_bias(vgg_layers, 34)) net['pool5'] = build_net('pool', net['conv5_4']) return netdef content_layer_loss(p, x): M = p.shape[1] * p.shape[2] N = p.shape[3] loss = (1. / (2 * N * M)) * tf.reduce_sum(tf.pow((x - p), 2)) return lossdef content_loss_func(sess, net): layers = CONTENT_LAYERS total_content_loss = 0.0 for layer_name, weight in layers: p = sess.run(net[layer_name]) x = net[layer_name] total_content_loss += content_layer_loss(p, x)*weight total_content_loss /= float(len(layers)) return total_content_lossdef gram_matrix(x, area, depth): x1 = tf.reshape(x, (area, depth)) g = tf.matmul(tf.transpose(x1), x1) return gdef style_layer_loss(a, x): M = a.shape[1] * a.shape[2] N = a.shape[3] A = gram_matrix(a, M, N) G = gram_matrix(x, M, N) loss = (1. / (4 * N ** 2 * M ** 2)) * tf.reduce_sum(tf.pow((G - A), 2)) return lossdef style_loss_func(sess, net): layers = STYLE_LAYERS total_style_loss = 0.0 for layer_name, weight in layers: a = sess.run(net[layer_name]) x = net[layer_name] total_style_loss += style_layer_loss(a, x) * weight total_style_loss /= float(len(layers)) return total_style_lossdef main(): net = build_vgg19(VGG_Model) sess = tf.Session() sess.run(tf.initialize_all_variables()) content_img = load_image(CONTENT_IMAGE) style_img = load_image(STYLE_IMAGE) sess.run([net['input'].assign(content_img)]) cost_content = content_loss_func(sess, net) sess.run([net['input'].assign(style_img)]) cost_style = style_loss_func(sess, net) total_loss = alpha * cost_content + beta * cost_style optimizer = tf.train.AdamOptimizer(2.0) init_img = generate_noise_image(content_img) train_op = optimizer.minimize(total_loss) sess.run(tf.initialize_all_variables()) sess.run(net['input'].assign(init_img)) for it in range(ITERATIONS): sess.run(train_op) if it % 100 == 0: # Print every 100 iteration. mixed_image = sess.run(net['input']) print('Iteration %d' % (it)) print('sum : ', sess.run(tf.reduce_sum(mixed_image))) print('cost: ', sess.run(total_loss)) if not os.path.exists(OUTPUT_DIR): os.mkdir(OUTPUT_DIR) filename = 'output/%d.png' % (it) save_image(filename, mixed_image)if __name__ == '__main__': main()
效果图