GAN生成式对抗生成网络

源代码：

# -*- coding = utf-8 -*-
# @Time : 2021/7/23
# @Author : pistachio
# @File : p26.py
# @Software : PyCharm

# GAN generator network
import keras
from keras import layers
import numpy as np
import os
from keras.preprocessing import image

latent_dim = 32
height = 32
width = 32
channels = 3

generator_input = keras.Input(shape=(latent_dim, ))

x = layers.Dense(128 * 16 * 16)(generator_input)
x = layers.LeakyReLU()(x)
x = layers.Reshape((16, 16, 128))(x)

x = layers.Conv2D(256, 5, padding='same')(x)
x = layers.LeakyReLU()(x)

x = layers.Conv2DTranspose(256, 4, strides=2, padding='same')(x)
x = layers.LeakyReLU()(x)

x = layers.Conv2D(256, 5, padding='same')(x)
x = layers.LeakyReLU()(x)
x = layers.Conv2D(256, 5, padding='same')(x)
x = layers.LeakyReLU()(x)

x = layers.Conv2D(channels, 7, activation='tanh', padding='same')(x)
generator = keras.models.Model(generator_input, x)
generator.summary()

#build GAN discriminator network
discriminator_input = layers.Input(shape=(height, width, channels))

x = layers.Conv2D(128, 3)(discriminator_input)
x = layers.LeakyReLU()(x)

x = layers.Conv2D(128, 4, strides=2)(x)
x = layers.LeakyReLU()(x)

x = layers.Conv2D(128, 4, strides=2)(x)
x = layers.LeakyReLU()(x)

x = layers.Conv2D(128, 4, strides=2)(x)
x = layers.LeakyReLU()(x)

x = layers.Flatten()(x)

x = layers.Dropout(0.4)(x)

x = layers.Dense(1, activation='sigmoid')(x)

discriminator = keras.models.Model(discriminator_input, x)
discriminator.summary()

discriminator_optimizer = keras.optimizers.RMSprop(
    lr=0.0008,
    clipvalue=1.0,
    decay=1e-8
)
discriminator.compile(
    optimizer=discriminator_optimizer,
    loss='binary_crossentropy'
)

discriminator.trainable = False
gan_input = keras.Input(shape=(latent_dim,))
gan_output = discriminator(generator(gan_input))
gan = keras.models.Model(gan_input, gan_output)
gan_optimizer = keras.optimizers.RMSprop(lr=0.0004, clipvalue=1.0, decay=1e-8)
gan.compile(optimizer=gan_optimizer, loss='binary_crossentropy')

#train GAN network
(x_train, y_train), (_, _) = keras.datasets.cifar10.load_data()
x_train = x_train[y_train.flatten() == 6]
x_train = x_train.reshape((x_train.shape[0],) +
                          (height, width, channels)).astype('float32') / 255.
iterations = 10000
batch_size = 20
save_dir = 'D:\PYCHARMprojects\Dailypractise\data\images'
start = 0

for step in range(iterations):
    random_latent_vectors = np.random.normal(size=(batch_size, latent_dim))
    generated_images = generator.predict(random_latent_vectors)
    
    stop = start + batch_size
    real_images = x_train[start: stop]
    combined_images = np.concatenate([generated_images, real_images])
    labels = np.concatenate([np.ones((batch_size, 1)),
                             np.zeros((batch_size, 1))])
    labels += 0.05 * np.random.random(labels.shape)
    d_loss = discriminator.test_on_batch(combined_images, labels)
    random_latent_vectors = np.random.normal(size=(batch_size, latent_dim))
    misleading_targets = np.zeros((batch_size, 1))
    a_loss = gan.train_on_batch(random_latent_vectors, misleading_targets)
    start += batch_size
    if start > len(x_train) - batch_size:
        start = 0
    
    if step % 100 == 0:
        gan.save_weights('gan.h5')
        
        print('discriminator loss:', d_loss)
        print('adversarial loss:', a_loss)
        img = image.array_to_img(generated_images[0] * 255., scale=False)
        img.save(os.path.join(save_dir, 'generated_frog' + str(step) + '.png'))
        img = image.array_to_img(real_images[0] * 255., scale=False)
        img.save(os.path.join(save_dir, 'real_frog' + str(step) + '.png'))

运行结果：

D:\Anaconda\envs\tensorflow\python.exe D:/PYCHARMprojects/Dailypractise/p26.py
2021-07-23 12:33:10.801058: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN)to use the following CPU instructions in performance-critical operations:  AVX AVX2
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
Model: "functional_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_1 (InputLayer)         [(None, 32)]              0         
_________________________________________________________________
dense (Dense)                (None, 32768)             1081344   
_________________________________________________________________
leaky_re_lu (LeakyReLU)      (None, 32768)             0         
_________________________________________________________________
reshape (Reshape)            (None, 16, 16, 128)       0         
_________________________________________________________________
conv2d (Conv2D)              (None, 16, 16, 256)       819456    
_________________________________________________________________
leaky_re_lu_1 (LeakyReLU)    (None, 16, 16, 256)       0         
_________________________________________________________________
conv2d_transpose (Conv2DTran (None, 32, 32, 256)       1048832   
_________________________________________________________________
leaky_re_lu_2 (LeakyReLU)    (None, 32, 32, 256)       0         
_________________________________________________________________
conv2d_1 (Conv2D)            (None, 32, 32, 256)       1638656   
_________________________________________________________________
leaky_re_lu_3 (LeakyReLU)    (None, 32, 32, 256)       0         
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 32, 32, 256)       1638656   
_________________________________________________________________
leaky_re_lu_4 (LeakyReLU)    (None, 32, 32, 256)       0         
_________________________________________________________________
conv2d_3 (Conv2D)            (None, 32, 32, 3)         37635     
=================================================================
Total params: 6,264,579
Trainable params: 6,264,579
Non-trainable params: 0
_________________________________________________________________
Model: "functional_3"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_2 (InputLayer)         [(None, 32, 32, 3)]       0         
_________________________________________________________________
conv2d_4 (Conv2D)            (None, 30, 30, 128)       3584      
_________________________________________________________________
leaky_re_lu_5 (LeakyReLU)    (None, 30, 30, 128)       0         
_________________________________________________________________
conv2d_5 (Conv2D)            (None, 14, 14, 128)       262272    
_________________________________________________________________
leaky_re_lu_6 (LeakyReLU)    (None, 14, 14, 128)       0         
_________________________________________________________________
conv2d_6 (Conv2D)            (None, 6, 6, 128)         262272    
_________________________________________________________________
leaky_re_lu_7 (LeakyReLU)    (None, 6, 6, 128)         0         
_________________________________________________________________
conv2d_7 (Conv2D)            (None, 2, 2, 128)         262272    
_________________________________________________________________
leaky_re_lu_8 (LeakyReLU)    (None, 2, 2, 128)         0         
_________________________________________________________________
flatten (Flatten)            (None, 512)               0         
_________________________________________________________________
dropout (Dropout)            (None, 512)               0         
_________________________________________________________________
dense_1 (Dense)              (None, 1)                 513       
=================================================================
Total params: 790,913
Trainable params: 790,913
Non-trainable params: 0
_________________________________________________________________
discriminator loss: 0.6984650492668152
adversarial loss: 0.6932023167610168

效果图：