In [ ]:
#@title Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
Note: 我们的 TensorFlow 社区翻译了这些文档。因为社区翻译是尽力而为, 所以无法保证它们是最准确的,并且反映了最新的 官方英文文档。如果您有改进此翻译的建议, 请提交 pull request 到 tensorflow/docs GitHub 仓库。要志愿地撰写或者审核译文,请加入 docs-zh-cn@tensorflow.org Google Group。
本教程演示了如何使用深度卷积生成对抗网络(DCGAN)生成手写数字图片。该代码是使用 Keras Sequential API 与 tf.GradientTape
训练循环编写的。
生成对抗网络(GANs)是当今计算机科学领域最有趣的想法之一。两个模型通过对抗过程同时训练。一个生成器(“艺术家”)学习创造看起来真实的图像,而判别器(“艺术评论家”)学习区分真假图像。
训练过程中,生成器在生成逼真图像方面逐渐变强,而判别器在辨别这些图像的能力上逐渐变强。当判别器不再能够区分真实图片和伪造图片时,训练过程达到平衡。
本笔记在 MNIST 数据集上演示了该过程。下方动画展示了当训练了 50 个epoch (全部数据集迭代50次) 时生成器所生成的一系列图片。图片从随机噪声开始,随着时间的推移越来越像手写数字。
要了解关于 GANs 的更多信息,我们建议参阅 MIT的 深度学习入门 课程。
In [ ]:
import tensorflow as tf
In [ ]:
tf.__version__
In [ ]:
# 用于生成 GIF 图片
!pip install imageio
In [ ]:
import glob
import imageio
import matplotlib.pyplot as plt
import numpy as np
import os
import PIL
from tensorflow.keras import layers
import time
from IPython import display
In [ ]:
(train_images, train_labels), (_, _) = tf.keras.datasets.mnist.load_data()
In [ ]:
train_images = train_images.reshape(train_images.shape[0], 28, 28, 1).astype('float32')
train_images = (train_images - 127.5) / 127.5 # 将图片标准化到 [-1, 1] 区间内
In [ ]:
BUFFER_SIZE = 60000
BATCH_SIZE = 256
In [ ]:
# 批量化和打乱数据
train_dataset = tf.data.Dataset.from_tensor_slices(train_images).shuffle(BUFFER_SIZE).batch(BATCH_SIZE)
生成器和判别器均使用 Keras Sequential API 定义。
In [ ]:
def make_generator_model():
model = tf.keras.Sequential()
model.add(layers.Dense(7*7*256, use_bias=False, input_shape=(100,)))
model.add(layers.BatchNormalization())
model.add(layers.LeakyReLU())
model.add(layers.Reshape((7, 7, 256)))
assert model.output_shape == (None, 7, 7, 256) # 注意:batch size 没有限制
model.add(layers.Conv2DTranspose(128, (5, 5), strides=(1, 1), padding='same', use_bias=False))
assert model.output_shape == (None, 7, 7, 128)
model.add(layers.BatchNormalization())
model.add(layers.LeakyReLU())
model.add(layers.Conv2DTranspose(64, (5, 5), strides=(2, 2), padding='same', use_bias=False))
assert model.output_shape == (None, 14, 14, 64)
model.add(layers.BatchNormalization())
model.add(layers.LeakyReLU())
model.add(layers.Conv2DTranspose(1, (5, 5), strides=(2, 2), padding='same', use_bias=False, activation='tanh'))
assert model.output_shape == (None, 28, 28, 1)
return model
使用(尚未训练的)生成器创建一张图片。
In [ ]:
generator = make_generator_model()
noise = tf.random.normal([1, 100])
generated_image = generator(noise, training=False)
plt.imshow(generated_image[0, :, :, 0], cmap='gray')
In [ ]:
def make_discriminator_model():
model = tf.keras.Sequential()
model.add(layers.Conv2D(64, (5, 5), strides=(2, 2), padding='same',
input_shape=[28, 28, 1]))
model.add(layers.LeakyReLU())
model.add(layers.Dropout(0.3))
model.add(layers.Conv2D(128, (5, 5), strides=(2, 2), padding='same'))
model.add(layers.LeakyReLU())
model.add(layers.Dropout(0.3))
model.add(layers.Flatten())
model.add(layers.Dense(1))
return model
使用(尚未训练的)判别器来对图片的真伪进行判断。模型将被训练为为真实图片输出正值,为伪造图片输出负值。
In [ ]:
discriminator = make_discriminator_model()
decision = discriminator(generated_image)
print (decision)
In [ ]:
# 该方法返回计算交叉熵损失的辅助函数
cross_entropy = tf.keras.losses.BinaryCrossentropy(from_logits=True)
In [ ]:
def discriminator_loss(real_output, fake_output):
real_loss = cross_entropy(tf.ones_like(real_output), real_output)
fake_loss = cross_entropy(tf.zeros_like(fake_output), fake_output)
total_loss = real_loss + fake_loss
return total_loss
In [ ]:
def generator_loss(fake_output):
return cross_entropy(tf.ones_like(fake_output), fake_output)
由于我们需要分别训练两个网络,判别器和生成器的优化器是不同的。
In [ ]:
generator_optimizer = tf.keras.optimizers.Adam(1e-4)
discriminator_optimizer = tf.keras.optimizers.Adam(1e-4)
In [ ]:
checkpoint_dir = './training_checkpoints'
checkpoint_prefix = os.path.join(checkpoint_dir, "ckpt")
checkpoint = tf.train.Checkpoint(generator_optimizer=generator_optimizer,
discriminator_optimizer=discriminator_optimizer,
generator=generator,
discriminator=discriminator)
In [ ]:
EPOCHS = 50
noise_dim = 100
num_examples_to_generate = 16
# 我们将重复使用该种子(因此在动画 GIF 中更容易可视化进度)
seed = tf.random.normal([num_examples_to_generate, noise_dim])
训练循环在生成器接收到一个随机种子作为输入时开始。该种子用于生产一张图片。判别器随后被用于区分真实图片(选自训练集)和伪造图片(由生成器生成)。针对这里的每一个模型都计算损失函数,并且计算梯度用于更新生成器与判别器。
In [ ]:
# 注意 `tf.function` 的使用
# 该注解使函数被“编译”
@tf.function
def train_step(images):
noise = tf.random.normal([BATCH_SIZE, noise_dim])
with tf.GradientTape() as gen_tape, tf.GradientTape() as disc_tape:
generated_images = generator(noise, training=True)
real_output = discriminator(images, training=True)
fake_output = discriminator(generated_images, training=True)
gen_loss = generator_loss(fake_output)
disc_loss = discriminator_loss(real_output, fake_output)
gradients_of_generator = gen_tape.gradient(gen_loss, generator.trainable_variables)
gradients_of_discriminator = disc_tape.gradient(disc_loss, discriminator.trainable_variables)
generator_optimizer.apply_gradients(zip(gradients_of_generator, generator.trainable_variables))
discriminator_optimizer.apply_gradients(zip(gradients_of_discriminator, discriminator.trainable_variables))
In [ ]:
def train(dataset, epochs):
for epoch in range(epochs):
start = time.time()
for image_batch in dataset:
train_step(image_batch)
# 继续进行时为 GIF 生成图像
display.clear_output(wait=True)
generate_and_save_images(generator,
epoch + 1,
seed)
# 每 15 个 epoch 保存一次模型
if (epoch + 1) % 15 == 0:
checkpoint.save(file_prefix = checkpoint_prefix)
print ('Time for epoch {} is {} sec'.format(epoch + 1, time.time()-start))
# 最后一个 epoch 结束后生成图片
display.clear_output(wait=True)
generate_and_save_images(generator,
epochs,
seed)
生成与保存图片
In [ ]:
def generate_and_save_images(model, epoch, test_input):
# 注意 training` 设定为 False
# 因此,所有层都在推理模式下运行(batchnorm)。
predictions = model(test_input, training=False)
fig = plt.figure(figsize=(4,4))
for i in range(predictions.shape[0]):
plt.subplot(4, 4, i+1)
plt.imshow(predictions[i, :, :, 0] * 127.5 + 127.5, cmap='gray')
plt.axis('off')
plt.savefig('image_at_epoch_{:04d}.png'.format(epoch))
plt.show()
In [ ]:
%%time
train(train_dataset, EPOCHS)
恢复最新的检查点。
In [ ]:
checkpoint.restore(tf.train.latest_checkpoint(checkpoint_dir))
In [ ]:
# 使用 epoch 数生成单张图片
def display_image(epoch_no):
return PIL.Image.open('image_at_epoch_{:04d}.png'.format(epoch_no))
In [ ]:
display_image(EPOCHS)
使用训练过程中生成的图片通过 imageio
生成动态 gif
In [ ]:
anim_file = 'dcgan.gif'
with imageio.get_writer(anim_file, mode='I') as writer:
filenames = glob.glob('image*.png')
filenames = sorted(filenames)
last = -1
for i,filename in enumerate(filenames):
frame = 2*(i**0.5)
if round(frame) > round(last):
last = frame
else:
continue
image = imageio.imread(filename)
writer.append_data(image)
image = imageio.imread(filename)
writer.append_data(image)
import IPython
if IPython.version_info > (6,2,0,''):
display.Image(filename=anim_file)
如果您正在使用 Colab,您可以通过如下代码下载动画:
In [ ]:
try:
from google.colab import files
except ImportError:
pass
else:
files.download(anim_file)
本教程展示了实现和训练 GAN 模型所需的全部必要代码。接下来,您可能想尝试其他数据集,例如大规模名人面部属性(CelebA)数据集 在 Kaggle 上获取。要了解更多关于 GANs 的信息,我们推荐参阅 NIPS 2016 教程: 生成对抗网络。