Denoising Autoencoder
Denoising Autoencoder
from tensorflow.keras.layers import Dense, Input, Conv2D, Flatten, Reshape, Conv2DTranspose
from tensorflow.keras.models import Model
from tensorflow.keras import backend as K
from tensorflow.keras.datasets import mnist
from tensorflow.keras.utils import plot_model
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from PIL import Image
np.random.seed(1234)
Dataset
(x_train, _), (x_test, _) = mnist.load_data()
image_size = x_train.shape[1]
x_train = np.reshape(x_train, [-1, image_size, image_size, 1])
x_test = np.reshape(x_test, [-1, image_size, image_size, 1])
x_train = x_train.astype('float32') / 255
x_test = x_test.astype('float32') / 255
add noise(mean = 0.5, std = 0.5)
noise = np.random.normal(loc=0.5, scale=0.5, size=x_train.shape)
x_train_noisy = x_train + noise
noise = np.random.normal(loc=0.5, scale=0.5, size=x_test.shape)
x_test_noisy = x_test + noise
fix [0, 1]
x_train_noisy = np.clip(x_train_noisy, 0., 1.)
x_test_noisy = np.clip(x_test_noisy, 0., 1.)
Hyperparameter
input_shape = (image_size, image_size, 1)
batch_size = 32
kernel_size = 3
latent_dim = 16
Model
Autoencoder
Encoder
inputs = Input(shape = input_shape, name='encoder_input')
x = inputs
x = Conv2D(filters=32, kernel_size=kernel_size, strides=2, activation='relu', padding='same')(x)
x = Conv2D(filters=64, kernel_size=kernel_size, strides=2, activation='relu', padding='same')(x)
shape = K.int_shape(x) # (7, 7, 64)
# Latent Vector
x = Flatten()(x)
latent = Dense(latent_dim, name='latent_vector')(x)
encoder = Model(inputs, latent, name='encoder')
encoder.summary()
Model: "encoder"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
encoder_input (InputLayer) [(None, 28, 28, 1)] 0
_________________________________________________________________
conv2d_4 (Conv2D) (None, 14, 14, 32) 320
_________________________________________________________________
conv2d_5 (Conv2D) (None, 7, 7, 64) 18496
_________________________________________________________________
flatten_1 (Flatten) (None, 3136) 0
_________________________________________________________________
latent_vector (Dense) (None, 16) 50192
=================================================================
Total params: 69,008
Trainable params: 69,008
Non-trainable params: 0
_________________________________________________________________
Decoder
latent_inputs = Input(shape=(latent_dim, ), name='decoder_input')
x = Dense(shape[1] * shape[2] * shape[3])(latent_inputs)
x = Reshape((shape[1], shape[2], shape[3]))(x)
x = Conv2DTranspose(filters=64, kernel_size=kernel_size, strides=2, activation='relu', padding='same')(x)
x = Conv2DTranspose(filters=32, kernel_size=kernel_size, strides=2, activation='relu', padding='same')(x)
outputs = Conv2DTranspose(filters=1, kernel_size=kernel_size, padding='same', activation='sigmoid',name='decoder_output')(x)
decoder = Model(latent_inputs, outputs, name='decoder')
decoder.summary()
Model: "decoder"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
decoder_input (InputLayer) [(None, 16)] 0
_________________________________________________________________
dense (Dense) (None, 3136) 53312
_________________________________________________________________
reshape (Reshape) (None, 7, 7, 64) 0
_________________________________________________________________
conv2d_transpose (Conv2DTran (None, 14, 14, 64) 36928
_________________________________________________________________
conv2d_transpose_1 (Conv2DTr (None, 28, 28, 32) 18464
_________________________________________________________________
decoder_output (Conv2DTransp (None, 28, 28, 1) 289
=================================================================
Total params: 108,993
Trainable params: 108,993
Non-trainable params: 0
_________________________________________________________________
AE = Model(inputs, decoder(encoder(inputs)), name='autoencoder')
AE.summary()
Model: "autoencoder"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
encoder_input (InputLayer) [(None, 28, 28, 1)] 0
_________________________________________________________________
encoder (Functional) (None, 16) 69008
_________________________________________________________________
decoder (Functional) (None, 28, 28, 1) 108993
=================================================================
Total params: 178,001
Trainable params: 178,001
Non-trainable params: 0
_________________________________________________________________
AE.compile(loss='mse', optimizer='adam')
AE.fit(x_train_noisy, x_train, validation_data=(x_test_noisy, x_test), epochs=10, batch_size=batch_size)
Epoch 1/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0365 - val_loss: 0.0206
Epoch 2/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0195 - val_loss: 0.0180
Epoch 3/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0179 - val_loss: 0.0176
Epoch 4/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0171 - val_loss: 0.0167
Epoch 5/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0165 - val_loss: 0.0164
Epoch 6/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0161 - val_loss: 0.0161
Epoch 7/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0157 - val_loss: 0.0159
Epoch 8/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0155 - val_loss: 0.0157
Epoch 9/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0153 - val_loss: 0.0157
Epoch 10/10
1875/1875 [==============================] - 5s 3ms/step - loss: 0.0151 - val_loss: 0.0157
<tensorflow.python.keras.callbacks.History at 0x1c395180b48>
x_decoded = AE.predict(x_test_noisy)
for i in range(0,4):
plt.figure(figsize=(10,4))
plt.subplot(1,3,1)
plt.imshow(x_test[i])
plt.subplot(1,3,2)
plt.imshow(x_test_noisy[i])
plt.subplot(1,3,3)
plt.imshow(x_decoded[i])
plt.show()
