-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathAAE_mnist.py
181 lines (147 loc) · 7.73 KB
/
AAE_mnist.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
import tensorflow as tf
import numpy as np
import os
import imageio
import matplotlib.pyplot as plt
from model import encoder, decoder, discriminator
class AAE_mnist():
def __init__(self,
n_dim=2,
batch_size=100,
epochs = 10,
log_freq=100,
results_path='./results',
make_gif=False):
self.n_dim = n_dim
self.batch_size = batch_size
self.epochs = epochs
self.log_freq = log_freq
self.results_path=results_path
self.results_img_path = results_path + "/imges"
self.make_gif = make_gif
if not os.path.exists(self.results_img_path):
os.makedirs(self.results_img_path)
if self.make_gif and not os.path.exists(self.results_path + "/gif"):
os.makedirs(self.results_path + "/gif")
# data load
self.load_data()
self.dataset_train = tf.data.Dataset.from_tensor_slices((self.x_train, self.y_train))
self.dtrain_shuffle = self.dataset_train.shuffle(self.x_train.shape[0]).batch(self.batch_size)
self.dataset_test = tf.data.Dataset.from_tensor_slices((self.x_test, self.y_test))
self.dtest_shuffle = self.dataset_test.shuffle(self.x_test.shape[0]).batch(1000)
# Models
self.encoder = encoder(n_dim=self.n_dim)
self.decoder = decoder()
self.discriminator = discriminator()
# optimizer
self.ae_opt = tf.keras.optimizers.Adam(0.0001)
self.gen_opt = tf.keras.optimizers.Adam(0.0001, beta_1=0, beta_2=0.9)
self.disc_opt = tf.keras.optimizers.Adam(0.0001, beta_1=0, beta_2=0.9)
self.loss_object = tf.keras.losses.BinaryCrossentropy(from_logits=True)
def load_data(self):
(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()
x_train = x_train.astype("float32") / 255.0
x_test = x_test.astype("float32") / 255.0
self.x_train = np.reshape(x_train, (-1, 28, 28, 1))
self.y_train = y_train
self.x_test = np.reshape(x_test, (-1, 28, 28, 1))
self.y_test = y_test
def generator_loss(self, discriminator_on_generator):
loss = self.loss_object(tf.ones_like(discriminator_on_generator), discriminator_on_generator)
return loss
def disc_real_loss(self, discriminator_on_data):
loss = self.loss_object(tf.ones_like(discriminator_on_data), discriminator_on_data)
return loss
def disc_fake_loss(self, discriminator_on_generator):
loss = self.loss_object(tf.zeros_like(discriminator_on_generator), discriminator_on_generator)
return loss
def train(self):
if self.make_gif:
self.z_sample = []
for i in np.linspace(-8, 8, 18):
for j in np.linspace(-8, 8, 18):
self.z_sample.append([i, j])
self.z_sample = np.asarray(self.z_sample)
# start training
for epoch in range(self.epochs):
print("\nStart of epoch : %d" % (epoch+1))
for step, (img_batch, label_batch) in enumerate(self.dtrain_shuffle):
# Autoencoder update
with tf.GradientTape() as tape:
z = self.encoder(img_batch)
recon_img = self.decoder(z)
recon_loss = tf.reduce_mean(tf.math.squared_difference(img_batch, recon_img))
grads = tape.gradient(recon_loss, self.encoder.trainable_weights + self.decoder.trainable_weights)
self.ae_opt.apply_gradients(zip(grads, self.encoder.trainable_weights + self.decoder.trainable_weights))
# Discriminator update
z = self.encoder(img_batch)
z_real = tf.random.normal([self.batch_size, self.n_dim], 0, 8)
with tf.GradientTape() as tape:
real_logits = self.discriminator(z_real)
fake_logits = self.discriminator(z)
fake_loss = self.disc_fake_loss(fake_logits)
real_loss = self.disc_real_loss(real_logits)
disc_loss = fake_loss + real_loss
grads = tape.gradient(disc_loss, self.discriminator.trainable_weights)
self.disc_opt.apply_gradients(zip(grads, self.discriminator.trainable_weights))
# Generator update
with tf.GradientTape() as tape:
z = self.encoder(img_batch)
gen_logits = self.discriminator(z)
gen_loss = self.generator_loss(gen_logits)
grads = tape.gradient(gen_loss, self.encoder.trainable_weights)
self.gen_opt.apply_gradients(zip(grads, self.encoder.trainable_weights))
if (step+1) % self.log_freq == 0:
print("epoch %d / %d, step %d / %d" % (epoch+1, self.epochs, step+1, self.x_train.shape[0]//self.batch_size))
print("\tgen_loss = %.4f" % (gen_loss))
print("\trecon_loss = %.4f" % (recon_loss))
print("\tdisc_loss = %.4f" % (disc_loss))
print("\t\tdisc_fake_loss = %.4f, disc_real_loss: %.4f" % (fake_loss, real_loss))
for test_img_batch, test_label in self.dtest_shuffle.take(1):
z = self.encoder(test_img_batch)
recon_img = self.decoder(z)
plt.scatter(z[:,0], z[:,1], s=0.5, c=test_label)
plt.colorbar()
plt.title("epoch: %d, step: %d" % (epoch+1, step+1))
plt.savefig(self.results_img_path + "/dist_" + str(epoch+1) + "_" + str(step+1))
plt.close()
# plt.show()
for i in range(16):
plt.subplot(4, 4, i + 1)
plt.imshow((recon_img[i]), cmap='Greys')
plt.axis('off')
plt.savefig(self.results_img_path + "/dist_" + str(epoch + 1) + "_" + str(step + 1))
plt.close()
# plt.show()
if self.make_gif:
z = self.encoder(self.x_test)
sample_img = self.decoder(self.z_sample)
plt.scatter(z[:,0], z[:,1], s=0.5, c=self.y_test)
plt.colorbar()
plt.title("epoch: %d" % (epoch + 1))
plt.xlim((-15, 15))
plt.ylim((-15, 15))
plt.savefig(self.results_path + "/gif/dist_" + str(epoch+1))
plt.close()
for i in range(18*18):
plt.subplot(18, 18, i+1)
plt.imshow((sample_img[i]), cmap='Greys')
plt.axis('off')
plt.savefig(self.results_path + "/gif/gen_" + str(epoch+1))
plt.close()
# save model
self.encoder.save(self.results_path + "/encoder")
self.decoder.save(self.results_path + "/decoder")
self.discriminator.save(self.results_path + "/discriminator")
if self.make_gif:
imgs_array = [np.array(imageio.imread(self.results_path + "/gif/dist_" + str(i+1) + ".png")) for i in range(self.epochs)]
imageio.mimsave(self.results_path + "/gif/dist.gif", imgs_array, duration=0.1)
imgs_array = [np.array(imageio.imread(self.results_path + "/gif/gen_" + str(i + 1) + ".png")) for i in range(self.epochs)]
imageio.mimsave(self.results_path + "/gif/gen.gif", imgs_array, duration=0.1)
if __name__ == "__main__":
AAE_mnist = AAE_mnist(n_dim=2,
epochs=50,
log_freq=600,
results_path='./results_mnist_AAE',
make_gif=True)
AAE_mnist.train()