GAN生成对抗网络-ACGAN原理与基本实现-条件生成对抗网络05

ACGAN介绍

案例一

import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers
import matplotlib.pyplot as plt
%matplotlib inline
import numpy as np
import glob

gpu = tf.config.experimental.list_physical_devices(device_type='GPU')
tf.config.experimental.set_memory_growth(gpu[0], True)

import tensorflow.keras.datasets.mnist as mnist

(train_image, train_label), (_, _) = mnist.load_data()

train_image = train_image / 127.5  - 1

train_image = np.expand_dims(train_image, -1)

dataset = tf.data.Dataset.from_tensor_slices((train_image, train_label))

AUTOTUNE = tf.data.experimental.AUTOTUNE

BATCH_SIZE = 256
image_count = train_image.shape[0]
noise_dim = 50

dataset = dataset.shuffle(image_count).batch(BATCH_SIZE)

def generator_model():
    seed = layers.Input(shape=((noise_dim,)))
    label = layers.Input(shape=(()))
    
    x = layers.Embedding(10, 50, input_length=1)(label)
    x = layers.Flatten()(x)
    x = layers.concatenate([seed, x])
    x = layers.Dense(3*3*128, use_bias=False)(x)
    x = layers.Reshape((3, 3, 128))(x)
    x = layers.BatchNormalization()(x)
    x = layers.ReLU()(x)
    
    x = layers.Conv2DTranspose(64, (3, 3), strides=(2, 2), use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.ReLU()(x)     #  7*7

    x = layers.Conv2DTranspose(32, (3, 3), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.ReLU()(x)    #   14*14

    x = layers.Conv2DTranspose(1, (3, 3), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.Activation('tanh')(x)
    
    model = tf.keras.Model(inputs=[seed,label], outputs=x)  
    
    return model

def discriminator_model():
    image = tf.keras.Input(shape=((28,28,1)))
    
    x = layers.Conv2D(32, (3, 3), strides=(2, 2), padding='same', use_bias=False)(image)
    x = layers.BatchNormalization()(x)
    x = layers.LeakyReLU()(x)
    x = layers.Dropout(0.5)(x)
    
    x = layers.Conv2D(32*2, (3, 3), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.LeakyReLU()(x)
    x = layers.Dropout(0.5)(x)
    
    x = layers.Conv2D(32*4, (3, 3), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.LeakyReLU()(x)
    x = layers.Dropout(0.5)(x)
    
    x = layers.Flatten()(x)
    x1 = layers.Dense(1)(x) # 真假输出
    x2 = layers.Dense(10)(x) # 分类输出
    
    model = tf.keras.Model(inputs=image, outputs=[x1, x2])
    return model

generator = generator_model()

discriminator = discriminator_model()

# 损失函数
binary_cross_entropy = tf.keras.losses.BinaryCrossentropy(from_logits=True) # 真假损失
category_cross_entropy = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True) # 交叉熵损失 多输出分类损失

def discriminator_loss(real_output, real_cat_out, fake_output, label): # 接收真图 和 真实图片的分类  假图 加label
    real_loss = binary_cross_entropy(tf.ones_like(real_output), real_output)
    fake_loss = binary_cross_entropy(tf.zeros_like(fake_output), fake_output)
    cat_loss = category_cross_entropy(label, real_cat_out)
    total_loss = real_loss + fake_loss + cat_loss
    return total_loss

def generator_loss(fake_output, fake_cat_out, label):
    fake_loss = binary_cross_entropy(tf.ones_like(fake_output), fake_output)
    cat_loss = category_cross_entropy(label, fake_cat_out)
    return fake_loss + cat_loss

generator_optimizer = tf.keras.optimizers.Adam(1e-5)
discriminator_optimizer = tf.keras.optimizers.Adam(1e-5)

@tf.function
def train_step(images, labels):
    batchsize = labels.shape[0]
    noise = tf.random.normal([batchsize, noise_dim])
    
    with tf.GradientTape() as gen_tape, tf.GradientTape() as disc_tape:
        generated_images = generator((noise, labels), training=True)

        real_output, real_cat_out = discriminator(images, training=True)
        fake_output, fake_cat_out = discriminator(generated_images, training=True)
        
        gen_loss = generator_loss(fake_output, fake_cat_out, labels)
        disc_loss = discriminator_loss(real_output, real_cat_out, fake_output, labels)

    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))

noise_dim = 50
num = 10
noise_seed = tf.random.normal([num, noise_dim])
cat_seed = np.random.randint(0, 10, size=(num, 1))
print(cat_seed.T)

def generate_and_save_images(model, test_noise_input, test_cat_input, epoch):
    print('Epoch:', epoch+1)
  # Notice `training` is set to False.
  # This is so all layers run in inference mode (batchnorm).
    predictions = model((test_noise_input, test_cat_input), training=False)
    predictions = tf.squeeze(predictions)
    fig = plt.figure(figsize=(10, 1))

    for i in range(predictions.shape[0]):
        plt.subplot(1, 10, i+1)
        plt.imshow((predictions[i, :, :] + 1)/2, cmap='gray')
        plt.axis('off')

#    plt.savefig('image_at_epoch_{:04d}.png'.format(epoch))
    plt.show()

def train(dataset, epochs):
    for epoch in range(epochs):
        for image_batch, label_batch in dataset:
            train_step(image_batch, label_batch)
        if epoch%10 == 0:
            generate_and_save_images(generator,
                                     noise_seed,
                                     cat_seed,
                                     epoch)


    generate_and_save_images(generator,
                            noise_seed,
                            cat_seed,
                            epoch)

EPOCHS = 200

train(dataset, EPOCHS)

generator.save('generate_acgan.h5')

num = 10
noise_seed = tf.random.normal([num, noise_dim])
cat_seed = np.arange(10).reshape(-1, 1)
print(cat_seed.T)

案例二

import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers
import matplotlib.pyplot as plt
%matplotlib inline
import numpy as np
import glob
import random

# 显存自适应分配
gpus = tf.config.experimental.list_physical_devices(device_type='GPU')
for gpu in gpus:
    tf.config.experimental.set_memory_growth(gpu,True)

gpu_ok = tf.test.is_gpu_available()
print("tf version:", tf.__version__)
print("use GPU", gpu_ok) # 判断是否使用gpu进行训练

with open('s.txt', 'w') as f:
    f.write('ddff')

import os

image_path = glob.glob("G:/BaiduNetdiskDownload/GAN生成对抗网络入门与实战/配套资料/face/*/*.jpg")

len(image_path)

random.seed(2020)
random.shuffle(image_path)

labels = [p.split("\\")[1] for p in image_path]

cls_to_num = dict((name, i) for i, name in enumerate(np.unique(labels)))

num_to_cls = {num: c for c, num in cls_to_num.items()}

labels = [cls_to_num.get(name) for name in labels]

image_path = np.array(image_path)
labels = np.array(labels)

# 编写图片加载函数预处理
@tf.function
def load_images(path):
    img = tf.io.read_file(path)
    img = tf.image.decode_jpeg(img, channels=3)# 解码
    img = tf.image.resize(img, [80, 80])# 把图片resize成80*80的大小
    img = tf.image.random_crop(img, [64, 64, 3])# 随机裁剪成64*64的大小 图像增强
    img = tf.image.random_flip_left_right(img)# 左右翻转
    img = tf.cast(img, tf.float32)/127.5 - 1# 归一化 255出127.5减1   0-1之间
    return img

img_dataset = tf.data.Dataset.from_tensor_slices(image_path)# 创建image的数据集

AUTOTUNE = tf.data.experimental.AUTOTUNE

img_dataset = img_dataset.map(load_images, num_parallel_calls=AUTOTUNE)

label_dataset = tf.data.Dataset.from_tensor_slices(labels)# 创建label数据集

dataset = tf.data.Dataset.zip((img_dataset, label_dataset))

BATCH_SIZE = 128
image_count = len(image_path)
noise_dim = 50

dataset = dataset.shuffle(300).batch(BATCH_SIZE)# 因为前面已经乱序我们这里只需要小范围乱序

def generator_model():
    seed = layers.Input(shape=((noise_dim,)))
    label = layers.Input(shape=(()))
    
    x = layers.Embedding(2, 50, input_length=1)(label)
    x = layers.Flatten()(x)
    x = layers.concatenate([seed, x])
    x = layers.Dense(4*4*64*8, use_bias=False)(x)
    x = layers.Reshape((4, 4, 64*8))(x)
    x = layers.BatchNormalization()(x)
    x = layers.ReLU()(x)
    
    x = layers.Conv2DTranspose(64*4, (5, 5), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.ReLU()(x)     #  8*8
    
    x = layers.Conv2DTranspose(64*2, (5, 5), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.ReLU()(x)    #   16*16 通道数64*2
    
    x = layers.Conv2DTranspose(64, (5, 5), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.ReLU()(x)    #    32*32
    
    x = layers.Conv2DTranspose(3, (5, 5), strides=(2, 2), padding='same', use_bias=False, activation='tanh')(x)
# 64*64*3
    model = tf.keras.Model(inputs=[seed,label], outputs=x)  
    
    return model

def discriminator_model():
    image = tf.keras.Input(shape=((64,64,3)))
    
    x = layers.Conv2D(64, (5, 5), strides=(2, 2), padding='same', use_bias=False)(image)
    x = layers.BatchNormalization()(x)
    x = layers.LeakyReLU()(x)

    x = layers.Conv2D(64*2, (5, 5), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.LeakyReLU()(x)

    x = layers.Conv2D(64*4, (5, 5), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.LeakyReLU()(x)

    x = layers.Conv2D(64*8, (5, 5), strides=(2, 2), padding='same', use_bias=False)(x)
    x = layers.BatchNormalization()(x)
    x = layers.LeakyReLU()(x)
    
    x = layers.Flatten()(x)
    x1 = layers.Dense(1)(x)# 真假输出
    x2 = layers.Dense(2)(x)# 分类输出
    
    model = tf.keras.Model(inputs=image, outputs=[x1, x2])
    return model

generator = generator_model()

discriminator = discriminator_model()

# 损失函数
binary_cross_entropy = tf.keras.losses.BinaryCrossentropy(from_logits=True)# 真假损失
category_cross_entropy = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)# 交叉熵损失 多输出分类损失

# 判别器损失
def discriminator_loss(real_output, real_cat_out, fake_output, label):
    real_loss = binary_cross_entropy(tf.ones_like(real_output), real_output)
    fake_loss = binary_cross_entropy(tf.zeros_like(fake_output), fake_output)
    cat_loss = category_cross_entropy(label, real_cat_out)
    total_loss = real_loss + fake_loss + cat_loss
    return total_loss

# 生成器损失
def generator_loss(fake_output, fake_cat_out, label):
    fake_loss = binary_cross_entropy(tf.ones_like(fake_output), fake_output)
    cat_loss = category_cross_entropy(label, fake_cat_out)
    return fake_loss + cat_loss

# 优化器
generator_optimizer = tf.keras.optimizers.Adam(1e-5)
discriminator_optimizer = tf.keras.optimizers.Adam(1e-5)

noise_dim = 50
num = 10
noise_seed = tf.random.normal([num, noise_dim])
cat_seed = np.random.randint(0, 2, size=(num, 1))
print(cat_seed.T)

condition = [' ' + num_to_cls.get(c) + '  ' for c in cat_seed.T[0]]
print(condition)

@tf.function
def train_step(images, labels):
    batchsize = labels.shape[0]
    noise = tf.random.normal([batchsize, noise_dim])
    
    with tf.GradientTape() as gen_tape, tf.GradientTape() as disc_tape:
        generated_images = generator((noise, labels), training=True)

        real_output, real_cat_out = discriminator(images, training=True)
        fake_output, fake_cat_out = discriminator(generated_images, training=True)
        
        gen_loss = generator_loss(fake_output, fake_cat_out, labels)
        disc_loss = discriminator_loss(real_output, real_cat_out, fake_output, labels)

    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))

def generate_and_save_images(model, test_noise_input, test_cat_input, epoch):
    print('Epoch:', epoch+1)
  # Notice `training` is set to False.
  # This is so all layers run in inference mode (batchnorm).
    predictions = model((test_noise_input, test_cat_input), training=False)

    fig = plt.figure(figsize=(20, 2))

    for i in range(predictions.shape[0]):
        plt.subplot(1, 10, i+1)
        plt.imshow((predictions[i, :, :, :] + 1)/2)
        plt.title(condition[i])
        plt.axis('off')

#    plt.savefig('image_at_epoch_{:04d}.png'.format(epoch))
    plt.show()

def train(dataset, epochs):
    for epoch in range(epochs):
        for image_batch, label_batch in dataset:
            train_step(image_batch, label_batch)
        if epoch%10 == 0:
            generate_and_save_images(generator,
                                     noise_seed,
                                     cat_seed,
                                     epoch)


    generate_and_save_images(generator,
                            noise_seed,
                            cat_seed,
                            epoch)

EPOCHS = 11000
train(dataset, EPOCHS)