DenseNet and Deconvolution Model Version 2
A Sparse-View CT Reconstruction Method Based on Combination of DenseNet and Deconvolution
UNet performed slightly better than this model.
I am trying to reimplement the paper below in keras
paper: https://ieeexplore.ieee.org/document/8331861 github: https://github.com/zzc623/DD_Net
from google.colab import drive
drive.mount('/content/drive')
Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly
Enter your authorization code:
··········
Mounted at /content/drive
paper: https://ieeexplore.ieee.org/document/8331861 github: https://github.com/zzc623/DD_Net
import os
import keras
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.image as mpimg
from keras.layers import Conv2D, MaxPooling2D, Input, Dense, UpSampling2D
from keras.layers import BatchNormalization, Activation, Dropout, Subtract
from keras.models import Model
from glob import glob
from keras.layers.convolutional import Conv2DTranspose
from keras.layers import concatenate
from keras import optimizers
Using TensorFlow backend.
The default version of TensorFlow in Colab will soon switch to TensorFlow 2.x.
We recommend you upgrade now
or ensure your notebook will continue to use TensorFlow 1.x via the %tensorflow_version 1.x magic:
more info.
# %tensorflow_version 1.x
# import tensorflow as tf
print(tf.__version__)
1.15.0
If we have weights we canload them: loaded_model.load_weights(“model.h5”)
# print(len(os.listdir('/content/drive/My Drive/Colab Notebooks/CT_data/sparseview_60/train')))
# print(len(os.listdir('/content/drive/My Drive/Colab Notebooks/CT_data/ndct/train')))
# # 3600
# # 3600
# print(len(os.listdir('/content/drive/My Drive/Colab Notebooks/CT_data/sparseview_60/test/')))
# print(len(os.listdir('/content/drive/My Drive/Colab Notebooks/CT_data/ndct/test/')))
# # # 354
# # # 354
ndct = sorted(glob('/content/drive/My Drive/Colab Notebooks/CT_data/ndct/train/*'))
ldct = sorted(glob('/content/drive/My Drive/Colab Notebooks/CT_data/sparseview_60/train/*'))
ndct_test = sorted(glob('/content/drive/My Drive/Colab Notebooks/CT_data/ndct/test/*'))
ldct_test = sorted(glob('/content/drive/My Drive/Colab Notebooks/CT_data/sparseview_60/test/*'))
print(len(ndct))
print(len(ldct))
print(len(ndct_test))
print(len(ldct_test))
3600
3600
354
354
The formulas below will be used to calculate the quality of the reconstruction. Higher PSNR generally indicates high quality of reconstruction.
def cal_psnr(im1, im2):
# assert pixel value range is 0-255 and type is uint8
mse = ((im1.astype(np.float) - im2.astype(np.float)) ** 2).mean()
maxval = np.amax(im1)
psnr = 10 * np.log10(maxval ** 2 / mse)
return psnr
def tf_psnr(im1, im2):
# assert pixel value range is 0-1
#mse = tf.losses.mean_squared_error(labels=im2 * 255.0, predictions=im1 * 255.0)
mse = tf.compat.v1.losses.mean_squared_error(labels=im2 * 255.0, predictions=im1 * 255.0)
return 10.0 * (tf.log(255.0 ** 2 / mse) / tf.log(10.0))
Using less data: #for i in range(0, 600). Have 3600 in all. Processing 3600 images takes aprox. 20 minutes to run. But once we create .npy aray we don’t have to rerun this code in the future and we will have .npy form of our data. Colab has 11GB RAM limit.
ndct_imgs_train = []
# for i in range(0, len(ndct)):
for i in range(0, 600):
f = open(ndct[i],'rb')
a = np.fromfile(f, np.float32)
ndct_imgs_train.append(a)
f.close()
print("len(ndct_imgs_train)....: ",len(ndct_imgs_train))
#len(ndct_imgs_train)....: 3600
Using different range to use less data to train #for i in range(0, 600). In all have 3600 images. It takes aprox. 20 minutes to process all 3600. But once we create .npy aray we don’t have to rerun this code in the future and we will have .npy form of our data.
ldct_imgs_train = []
# for i in range(0, len(ldct)):
for i in range(0, 600):
f = open(ldct[i],'rb')
a = np.fromfile(f, np.float32)
ldct_imgs_train.append(a)
f.close()
print("len(ldct_imgs_train)....: ",len(ldct_imgs_train))
only using 100 images to test
ndct_imgs_test = []
# for i in range(0, len(ndct_test)):
for i in range(0, 100):
f = open(ndct_test[i],'rb')
a = np.fromfile(f, np.float32)
ndct_imgs_test.append(a)
f.close()
print("len(ndct_imgs_test)....: ",len(ndct_imgs_test))
only using 100 images to test
# load the image
ldct_imgs_test = []
# for i in range(0, len(ldct_test)):
for i in range(0, 100):
f = open(ldct_test[i],'rb')
a = np.fromfile(f, np.float32)
ldct_imgs_test.append(a)
f.close()
print("len(ldct_imgs_test)....: ",len(ldct_imgs_test))
Must reshape images to train
ldct_train = np.asarray(ldct_imgs_train)
ndct_train = np.asarray(ndct_imgs_train)
ldct_train = ldct_train.reshape(600,512,512,1)
ndct_train = ndct_train.reshape(600,512,512,1)
ldct_test = np.asarray(ldct_imgs_test)
ndct_test = np.asarray(ndct_imgs_test)
# ldct_test = ldct_test.reshape(len(ldct_imgs_test),512,512,1)
# ndct_test = ndct_test.reshape(len(ldct_imgs_test),512,512,1)
ldct_test = ldct_test.reshape(100,512,512,1)
ndct_test = ndct_test.reshape(100,512,512,1)
print(ldct_train.shape)
print(ndct_train.shape)
print(ldct_test.shape)
print(ndct_test.shape)
# np.save('sparseview_60_train_600', ldct_train) # save the file as "sparseview_60_train.npy"
# np.save('ndct_train_600', ndct_train) # save the file as "ndct_train.npy"
# np.save('sparseview_60_test_100', ldct_test) # save the file as "sparseview_60_test.npy"
# np.save('ndct_test_100', ndct_test) # save the file as "ndct_test.npy"
np.save('/content/drive/My Drive/Colab Notebooks/dd_net/sparseview_60_train_600', ldct_train) # save the file as "sparseview_60_train.npy"
np.save('/content/drive/My Drive/Colab Notebooks/dd_net/ndct_train_600', ndct_train) # save the file as "ndct_train.npy"
np.save('/content/drive/My Drive/Colab Notebooks/dd_net/sparseview_60_test_100', ldct_test) # save the file as "sparseview_60_test.npy"
np.save('/content/drive/My Drive/Colab Notebooks/dd_net/ndct_test_100', ndct_test) # save the file as "ndct_test.npy"
sparseview_60_train = np.load('/content/drive/My Drive/Colab Notebooks/dd_net/sparseview_60_train_600.npy') # loads saved array into variable sparseview_60_train.
ndct_train = np.load('/content/drive/My Drive/Colab Notebooks/dd_net/ndct_train_600.npy') # loads saved array into variable ndct_train.
sparseview_60_test = np.load('/content/drive/My Drive/Colab Notebooks/dd_net/sparseview_60_test_100.npy') # loads saved array into variable sparseview_60_test.
ndct_test = np.load('/content/drive/My Drive/Colab Notebooks/dd_net/ndct_test_100.npy') # loads saved array into variable ndct_test.
# sparseview_60_train = np.load('sparseview_60_train_600.npy') # loads saved array into variable sparseview_60_train.
# ndct_train = np.load('ndct_train_600.npy') # loads saved array into variable ndct_train.
# sparseview_60_test = np.load('sparseview_60_test_100.npy') # loads saved array into variable sparseview_60_test.
# ndct_test = np.load('ndct_test_100.npy') # loads saved array into variable ndct_test.
def denseblock(input):
# - L1
num_filters = 16
d2_1 = BatchNormalization()(input)
d2_1 = Activation('relu')(d2_1)
d2_1 = Conv2D(num_filters*4, (1, 1), padding='same', use_bias=True, strides=(1, 1))(d2_1)
d2_1 = BatchNormalization()(d2_1)
d2_1 = Activation('relu')(d2_1)
d2_1 = Conv2D(num_filters, (5, 5), padding='same', use_bias=True, strides=(1, 1))(d2_1)
d2_1 = concatenate([input, d2_1])
# - L2
d2_2 = BatchNormalization()(d2_1)
d2_2 = Activation('relu')(d2_2)
d2_2 = Conv2D(num_filters*4, (1, 1), padding='same', use_bias=True, strides=(1, 1))(d2_2)
d2_2 = BatchNormalization()(d2_2)
d2_2 = Activation('relu')(d2_2)
d2_2 = Conv2D(num_filters, (5, 5), padding='same', use_bias=True, strides=(1, 1))(d2_2)
d2_2 = concatenate([input, d2_1, d2_2])
# - L3
d2_3 = BatchNormalization()(d2_2)
d2_3 = Activation('relu')(d2_3)
d2_3 = Conv2D(num_filters*4, (1, 1), padding='same', use_bias=True, strides=(1, 1))(d2_3)
d2_3 = BatchNormalization()(d2_3)
d2_3 = Activation('relu')(d2_3)
d2_3 = Conv2D(num_filters, (5, 5), padding='same', use_bias=True, strides=(1, 1))(d2_3)
d2_3 = concatenate([input, d2_1, d2_2, d2_3])
# - L4
d2_4 = BatchNormalization()(d2_3)
d2_4 = Activation('relu')(d2_4)
d2_4 = Conv2D(num_filters*4, (1, 1), padding='same', use_bias=True, strides=(1, 1))(d2_4)
d2_4 = BatchNormalization()(d2_4)
d2_4 = Activation('relu')(d2_4)
d2_4 = Conv2D(num_filters, (5, 5), padding='same', use_bias=True, strides=(1, 1))(d2_4)
d2_4 = concatenate([input, d2_1, d2_2, d2_3, d2_4])
return d2_4
inputs = Input((None, None,1))
num_filter = 16
# ---A1 Layer-----------------------
h_conv1 = Conv2D(16, (7, 7), padding='same', use_bias=True, strides=(1, 1))(inputs)
a1 = MaxPooling2D((3, 3), strides=(2, 2), padding='same') (h_conv1)
# images 256 X 256
d1 = denseblock(a1)
a1 = BatchNormalization()(d1)
a1 = Activation('relu')(a1)
h_conv1_T = Conv2D(16, (1, 1), strides=(1, 1), use_bias=True) (a1)
# ----A2 Layer---------------------
a2 = MaxPooling2D((2, 2),strides=(2, 2), padding='same') (h_conv1_T)
# images 128 X 128 d
d2 = denseblock(a2)
a2 = BatchNormalization()(d2)
a2 = Activation('relu')(a2)
h_conv2_T = Conv2D(16, (1, 1), strides=(1, 1), use_bias=True) (a2)
# images 128 X 128
# # ----A3 Layer----------------------
a3 = MaxPooling2D((2, 2), strides=(2, 2), padding='same') (h_conv2_T)
# images 64 X 64
d3 = denseblock(a3)
a3 = BatchNormalization()(d3)
a3 = Activation('relu')(a3)
h_conv3_T = Conv2D(16, (1, 1), strides=(1, 1), use_bias=True) (a3)
# ----A4 Layer----------------------
a4 = MaxPooling2D((2, 2), strides=(2, 2), padding='same') (h_conv3_T)
# images 32 X 3
d4 = denseblock(a4)
a4 = BatchNormalization()(d4)
a4 = Activation('relu')(a4)
h_conv4_T = Conv2D(16, (1, 1), strides=(1, 1), use_bias=True) (a4)
# #----B1 Layer-----------------------
b1 = UpSampling2D((2, 2), interpolation="nearest") (h_conv4_T)
# images 64 X 64
b1 = concatenate([b1, h_conv3_T])
b1 = Conv2DTranspose(num_filter*2, (5, 5), padding='same', strides=(1, 1)) (b1)
b1 = Activation('relu')(b1)
b1 = BatchNormalization()(b1)
b1 = Conv2DTranspose(16, (1, 1), padding='same', strides=(1, 1)) (b1)
b1 = Activation('relu')(b1)
b1 = BatchNormalization()(b1)
# #----B2 Layer-----------------------
b2 = UpSampling2D((2, 2), interpolation="nearest") (b1)
# images 128 X 128
b2 = concatenate([b2, h_conv2_T])
b2 = Conv2DTranspose(num_filter*2, (5, 5), padding='same', strides=(1, 1)) (b2)
b2 = Activation('relu')(b2)
b2 = BatchNormalization()(b2)
b2 = Conv2DTranspose(16, (1, 1), padding='same', strides=(1, 1)) (b2)
b2 = Activation('relu')(b2)
b2 = BatchNormalization()(b2)
#----B3 Layer------------------------conv6
b3 = UpSampling2D((2, 2),interpolation="nearest") (b2)
# images 256 X 256
b3 = concatenate([b3, h_conv1_T])
b3 = Conv2DTranspose(num_filter*2, (5, 5), padding='same', strides=(1, 1)) (b3)
b3 = Activation('relu')(b3)
b3 = BatchNormalization()(b3)
b3 = Conv2DTranspose(16, (1, 1), padding='same', strides=(1, 1)) (b3)
b3 = Activation('relu')(b3)
b3 = BatchNormalization()(b3)
#----B4 Layer-------------------------
b4 = UpSampling2D((2, 2),interpolation="nearest") (b3)
# images 512 X 512
b4 = concatenate([b4, h_conv1])
b4 = Conv2DTranspose(num_filter*2, (5, 5),padding='same', strides=(1, 1)) (b4)
b4 = Activation('relu')(b4)
# b4 = BatchNormalization()(b4)
output_img = Conv2DTranspose(1, (1, 1), strides=(1, 1)) (b4)
# output_img = Activation('relu')(output_img) # in paper but DIDN'T CONVERGE
# ------ end B4 layer
subtracted = Subtract()([inputs, output_img])
ddnet_model = Model(inputs=[inputs], outputs=[subtracted])
ddnet_model.compile(optimizer='adam', loss='mse', metrics=[tf_psnr])
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:66: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:541: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4432: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4267: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:190: The name tf.get_default_session is deprecated. Please use tf.compat.v1.get_default_session instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:197: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:203: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:207: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:216: The name tf.is_variable_initialized is deprecated. Please use tf.compat.v1.is_variable_initialized instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:223: The name tf.variables_initializer is deprecated. Please use tf.compat.v1.variables_initializer instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:2041: The name tf.nn.fused_batch_norm is deprecated. Please use tf.compat.v1.nn.fused_batch_norm instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:148: The name tf.placeholder_with_default is deprecated. Please use tf.compat.v1.placeholder_with_default instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:2239: The name tf.image.resize_nearest_neighbor is deprecated. Please use tf.compat.v1.image.resize_nearest_neighbor instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/optimizers.py:793: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/losses/losses_impl.py:121: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use tf.where in 2.0, which has the same broadcast rule as np.where
A Sparse-View CT Reconstruction Method Based on Combination of DenseNet and Deconvolution
ddnet_model.summary()
Model: "model_1"
__________________________________________________________________________________________________
Layer (type) Output Shape Param # Connected to
==================================================================================================
input_1 (InputLayer) (None, None, None, 1 0
__________________________________________________________________________________________________
conv2d_1 (Conv2D) (None, None, None, 1 800 input_1[0][0]
__________________________________________________________________________________________________
max_pooling2d_1 (MaxPooling2D) (None, None, None, 1 0 conv2d_1[0][0]
__________________________________________________________________________________________________
batch_normalization_1 (BatchNor (None, None, None, 1 64 max_pooling2d_1[0][0]
__________________________________________________________________________________________________
activation_1 (Activation) (None, None, None, 1 0 batch_normalization_1[0][0]
__________________________________________________________________________________________________
conv2d_2 (Conv2D) (None, None, None, 6 1088 activation_1[0][0]
__________________________________________________________________________________________________
batch_normalization_2 (BatchNor (None, None, None, 6 256 conv2d_2[0][0]
__________________________________________________________________________________________________
activation_2 (Activation) (None, None, None, 6 0 batch_normalization_2[0][0]
__________________________________________________________________________________________________
conv2d_3 (Conv2D) (None, None, None, 1 25616 activation_2[0][0]
__________________________________________________________________________________________________
concatenate_1 (Concatenate) (None, None, None, 3 0 max_pooling2d_1[0][0]
conv2d_3[0][0]
__________________________________________________________________________________________________
batch_normalization_3 (BatchNor (None, None, None, 3 128 concatenate_1[0][0]
__________________________________________________________________________________________________
activation_3 (Activation) (None, None, None, 3 0 batch_normalization_3[0][0]
__________________________________________________________________________________________________
conv2d_4 (Conv2D) (None, None, None, 6 2112 activation_3[0][0]
__________________________________________________________________________________________________
batch_normalization_4 (BatchNor (None, None, None, 6 256 conv2d_4[0][0]
__________________________________________________________________________________________________
activation_4 (Activation) (None, None, None, 6 0 batch_normalization_4[0][0]
__________________________________________________________________________________________________
conv2d_5 (Conv2D) (None, None, None, 1 25616 activation_4[0][0]
__________________________________________________________________________________________________
concatenate_2 (Concatenate) (None, None, None, 6 0 max_pooling2d_1[0][0]
concatenate_1[0][0]
conv2d_5[0][0]
__________________________________________________________________________________________________
batch_normalization_5 (BatchNor (None, None, None, 6 256 concatenate_2[0][0]
__________________________________________________________________________________________________
activation_5 (Activation) (None, None, None, 6 0 batch_normalization_5[0][0]
__________________________________________________________________________________________________
conv2d_6 (Conv2D) (None, None, None, 6 4160 activation_5[0][0]
__________________________________________________________________________________________________
batch_normalization_6 (BatchNor (None, None, None, 6 256 conv2d_6[0][0]
__________________________________________________________________________________________________
activation_6 (Activation) (None, None, None, 6 0 batch_normalization_6[0][0]
__________________________________________________________________________________________________
conv2d_7 (Conv2D) (None, None, None, 1 25616 activation_6[0][0]
__________________________________________________________________________________________________
concatenate_3 (Concatenate) (None, None, None, 1 0 max_pooling2d_1[0][0]
concatenate_1[0][0]
concatenate_2[0][0]
conv2d_7[0][0]
__________________________________________________________________________________________________
batch_normalization_7 (BatchNor (None, None, None, 1 512 concatenate_3[0][0]
__________________________________________________________________________________________________
activation_7 (Activation) (None, None, None, 1 0 batch_normalization_7[0][0]
__________________________________________________________________________________________________
conv2d_8 (Conv2D) (None, None, None, 6 8256 activation_7[0][0]
__________________________________________________________________________________________________
batch_normalization_8 (BatchNor (None, None, None, 6 256 conv2d_8[0][0]
__________________________________________________________________________________________________
activation_8 (Activation) (None, None, None, 6 0 batch_normalization_8[0][0]
__________________________________________________________________________________________________
conv2d_9 (Conv2D) (None, None, None, 1 25616 activation_8[0][0]
__________________________________________________________________________________________________
concatenate_4 (Concatenate) (None, None, None, 2 0 max_pooling2d_1[0][0]
concatenate_1[0][0]
concatenate_2[0][0]
concatenate_3[0][0]
conv2d_9[0][0]
__________________________________________________________________________________________________
batch_normalization_9 (BatchNor (None, None, None, 2 1024 concatenate_4[0][0]
__________________________________________________________________________________________________
activation_9 (Activation) (None, None, None, 2 0 batch_normalization_9[0][0]
__________________________________________________________________________________________________
conv2d_10 (Conv2D) (None, None, None, 1 4112 activation_9[0][0]
__________________________________________________________________________________________________
max_pooling2d_2 (MaxPooling2D) (None, None, None, 1 0 conv2d_10[0][0]
__________________________________________________________________________________________________
batch_normalization_10 (BatchNo (None, None, None, 1 64 max_pooling2d_2[0][0]
__________________________________________________________________________________________________
activation_10 (Activation) (None, None, None, 1 0 batch_normalization_10[0][0]
__________________________________________________________________________________________________
conv2d_11 (Conv2D) (None, None, None, 6 1088 activation_10[0][0]
__________________________________________________________________________________________________
batch_normalization_11 (BatchNo (None, None, None, 6 256 conv2d_11[0][0]
__________________________________________________________________________________________________
activation_11 (Activation) (None, None, None, 6 0 batch_normalization_11[0][0]
__________________________________________________________________________________________________
conv2d_12 (Conv2D) (None, None, None, 1 25616 activation_11[0][0]
__________________________________________________________________________________________________
concatenate_5 (Concatenate) (None, None, None, 3 0 max_pooling2d_2[0][0]
conv2d_12[0][0]
__________________________________________________________________________________________________
batch_normalization_12 (BatchNo (None, None, None, 3 128 concatenate_5[0][0]
__________________________________________________________________________________________________
activation_12 (Activation) (None, None, None, 3 0 batch_normalization_12[0][0]
__________________________________________________________________________________________________
conv2d_13 (Conv2D) (None, None, None, 6 2112 activation_12[0][0]
__________________________________________________________________________________________________
batch_normalization_13 (BatchNo (None, None, None, 6 256 conv2d_13[0][0]
__________________________________________________________________________________________________
activation_13 (Activation) (None, None, None, 6 0 batch_normalization_13[0][0]
__________________________________________________________________________________________________
conv2d_14 (Conv2D) (None, None, None, 1 25616 activation_13[0][0]
__________________________________________________________________________________________________
concatenate_6 (Concatenate) (None, None, None, 6 0 max_pooling2d_2[0][0]
concatenate_5[0][0]
conv2d_14[0][0]
__________________________________________________________________________________________________
batch_normalization_14 (BatchNo (None, None, None, 6 256 concatenate_6[0][0]
__________________________________________________________________________________________________
activation_14 (Activation) (None, None, None, 6 0 batch_normalization_14[0][0]
__________________________________________________________________________________________________
conv2d_15 (Conv2D) (None, None, None, 6 4160 activation_14[0][0]
__________________________________________________________________________________________________
batch_normalization_15 (BatchNo (None, None, None, 6 256 conv2d_15[0][0]
__________________________________________________________________________________________________
activation_15 (Activation) (None, None, None, 6 0 batch_normalization_15[0][0]
__________________________________________________________________________________________________
conv2d_16 (Conv2D) (None, None, None, 1 25616 activation_15[0][0]
__________________________________________________________________________________________________
concatenate_7 (Concatenate) (None, None, None, 1 0 max_pooling2d_2[0][0]
concatenate_5[0][0]
concatenate_6[0][0]
conv2d_16[0][0]
__________________________________________________________________________________________________
batch_normalization_16 (BatchNo (None, None, None, 1 512 concatenate_7[0][0]
__________________________________________________________________________________________________
activation_16 (Activation) (None, None, None, 1 0 batch_normalization_16[0][0]
__________________________________________________________________________________________________
conv2d_17 (Conv2D) (None, None, None, 6 8256 activation_16[0][0]
__________________________________________________________________________________________________
batch_normalization_17 (BatchNo (None, None, None, 6 256 conv2d_17[0][0]
__________________________________________________________________________________________________
activation_17 (Activation) (None, None, None, 6 0 batch_normalization_17[0][0]
__________________________________________________________________________________________________
conv2d_18 (Conv2D) (None, None, None, 1 25616 activation_17[0][0]
__________________________________________________________________________________________________
concatenate_8 (Concatenate) (None, None, None, 2 0 max_pooling2d_2[0][0]
concatenate_5[0][0]
concatenate_6[0][0]
concatenate_7[0][0]
conv2d_18[0][0]
__________________________________________________________________________________________________
batch_normalization_18 (BatchNo (None, None, None, 2 1024 concatenate_8[0][0]
__________________________________________________________________________________________________
activation_18 (Activation) (None, None, None, 2 0 batch_normalization_18[0][0]
__________________________________________________________________________________________________
conv2d_19 (Conv2D) (None, None, None, 1 4112 activation_18[0][0]
__________________________________________________________________________________________________
max_pooling2d_3 (MaxPooling2D) (None, None, None, 1 0 conv2d_19[0][0]
__________________________________________________________________________________________________
batch_normalization_19 (BatchNo (None, None, None, 1 64 max_pooling2d_3[0][0]
__________________________________________________________________________________________________
activation_19 (Activation) (None, None, None, 1 0 batch_normalization_19[0][0]
__________________________________________________________________________________________________
conv2d_20 (Conv2D) (None, None, None, 6 1088 activation_19[0][0]
__________________________________________________________________________________________________
batch_normalization_20 (BatchNo (None, None, None, 6 256 conv2d_20[0][0]
__________________________________________________________________________________________________
activation_20 (Activation) (None, None, None, 6 0 batch_normalization_20[0][0]
__________________________________________________________________________________________________
conv2d_21 (Conv2D) (None, None, None, 1 25616 activation_20[0][0]
__________________________________________________________________________________________________
concatenate_9 (Concatenate) (None, None, None, 3 0 max_pooling2d_3[0][0]
conv2d_21[0][0]
__________________________________________________________________________________________________
batch_normalization_21 (BatchNo (None, None, None, 3 128 concatenate_9[0][0]
__________________________________________________________________________________________________
activation_21 (Activation) (None, None, None, 3 0 batch_normalization_21[0][0]
__________________________________________________________________________________________________
conv2d_22 (Conv2D) (None, None, None, 6 2112 activation_21[0][0]
__________________________________________________________________________________________________
batch_normalization_22 (BatchNo (None, None, None, 6 256 conv2d_22[0][0]
__________________________________________________________________________________________________
activation_22 (Activation) (None, None, None, 6 0 batch_normalization_22[0][0]
__________________________________________________________________________________________________
conv2d_23 (Conv2D) (None, None, None, 1 25616 activation_22[0][0]
__________________________________________________________________________________________________
concatenate_10 (Concatenate) (None, None, None, 6 0 max_pooling2d_3[0][0]
concatenate_9[0][0]
conv2d_23[0][0]
__________________________________________________________________________________________________
batch_normalization_23 (BatchNo (None, None, None, 6 256 concatenate_10[0][0]
__________________________________________________________________________________________________
activation_23 (Activation) (None, None, None, 6 0 batch_normalization_23[0][0]
__________________________________________________________________________________________________
conv2d_24 (Conv2D) (None, None, None, 6 4160 activation_23[0][0]
__________________________________________________________________________________________________
batch_normalization_24 (BatchNo (None, None, None, 6 256 conv2d_24[0][0]
__________________________________________________________________________________________________
activation_24 (Activation) (None, None, None, 6 0 batch_normalization_24[0][0]
__________________________________________________________________________________________________
conv2d_25 (Conv2D) (None, None, None, 1 25616 activation_24[0][0]
__________________________________________________________________________________________________
concatenate_11 (Concatenate) (None, None, None, 1 0 max_pooling2d_3[0][0]
concatenate_9[0][0]
concatenate_10[0][0]
conv2d_25[0][0]
__________________________________________________________________________________________________
batch_normalization_25 (BatchNo (None, None, None, 1 512 concatenate_11[0][0]
__________________________________________________________________________________________________
activation_25 (Activation) (None, None, None, 1 0 batch_normalization_25[0][0]
__________________________________________________________________________________________________
conv2d_26 (Conv2D) (None, None, None, 6 8256 activation_25[0][0]
__________________________________________________________________________________________________
batch_normalization_26 (BatchNo (None, None, None, 6 256 conv2d_26[0][0]
__________________________________________________________________________________________________
activation_26 (Activation) (None, None, None, 6 0 batch_normalization_26[0][0]
__________________________________________________________________________________________________
conv2d_27 (Conv2D) (None, None, None, 1 25616 activation_26[0][0]
__________________________________________________________________________________________________
concatenate_12 (Concatenate) (None, None, None, 2 0 max_pooling2d_3[0][0]
concatenate_9[0][0]
concatenate_10[0][0]
concatenate_11[0][0]
conv2d_27[0][0]
__________________________________________________________________________________________________
batch_normalization_27 (BatchNo (None, None, None, 2 1024 concatenate_12[0][0]
__________________________________________________________________________________________________
activation_27 (Activation) (None, None, None, 2 0 batch_normalization_27[0][0]
__________________________________________________________________________________________________
conv2d_28 (Conv2D) (None, None, None, 1 4112 activation_27[0][0]
__________________________________________________________________________________________________
max_pooling2d_4 (MaxPooling2D) (None, None, None, 1 0 conv2d_28[0][0]
__________________________________________________________________________________________________
batch_normalization_28 (BatchNo (None, None, None, 1 64 max_pooling2d_4[0][0]
__________________________________________________________________________________________________
activation_28 (Activation) (None, None, None, 1 0 batch_normalization_28[0][0]
__________________________________________________________________________________________________
conv2d_29 (Conv2D) (None, None, None, 6 1088 activation_28[0][0]
__________________________________________________________________________________________________
batch_normalization_29 (BatchNo (None, None, None, 6 256 conv2d_29[0][0]
__________________________________________________________________________________________________
activation_29 (Activation) (None, None, None, 6 0 batch_normalization_29[0][0]
__________________________________________________________________________________________________
conv2d_30 (Conv2D) (None, None, None, 1 25616 activation_29[0][0]
__________________________________________________________________________________________________
concatenate_13 (Concatenate) (None, None, None, 3 0 max_pooling2d_4[0][0]
conv2d_30[0][0]
__________________________________________________________________________________________________
batch_normalization_30 (BatchNo (None, None, None, 3 128 concatenate_13[0][0]
__________________________________________________________________________________________________
activation_30 (Activation) (None, None, None, 3 0 batch_normalization_30[0][0]
__________________________________________________________________________________________________
conv2d_31 (Conv2D) (None, None, None, 6 2112 activation_30[0][0]
__________________________________________________________________________________________________
batch_normalization_31 (BatchNo (None, None, None, 6 256 conv2d_31[0][0]
__________________________________________________________________________________________________
activation_31 (Activation) (None, None, None, 6 0 batch_normalization_31[0][0]
__________________________________________________________________________________________________
conv2d_32 (Conv2D) (None, None, None, 1 25616 activation_31[0][0]
__________________________________________________________________________________________________
concatenate_14 (Concatenate) (None, None, None, 6 0 max_pooling2d_4[0][0]
concatenate_13[0][0]
conv2d_32[0][0]
__________________________________________________________________________________________________
batch_normalization_32 (BatchNo (None, None, None, 6 256 concatenate_14[0][0]
__________________________________________________________________________________________________
activation_32 (Activation) (None, None, None, 6 0 batch_normalization_32[0][0]
__________________________________________________________________________________________________
conv2d_33 (Conv2D) (None, None, None, 6 4160 activation_32[0][0]
__________________________________________________________________________________________________
batch_normalization_33 (BatchNo (None, None, None, 6 256 conv2d_33[0][0]
__________________________________________________________________________________________________
activation_33 (Activation) (None, None, None, 6 0 batch_normalization_33[0][0]
__________________________________________________________________________________________________
conv2d_34 (Conv2D) (None, None, None, 1 25616 activation_33[0][0]
__________________________________________________________________________________________________
concatenate_15 (Concatenate) (None, None, None, 1 0 max_pooling2d_4[0][0]
concatenate_13[0][0]
concatenate_14[0][0]
conv2d_34[0][0]
__________________________________________________________________________________________________
batch_normalization_34 (BatchNo (None, None, None, 1 512 concatenate_15[0][0]
__________________________________________________________________________________________________
activation_34 (Activation) (None, None, None, 1 0 batch_normalization_34[0][0]
__________________________________________________________________________________________________
conv2d_35 (Conv2D) (None, None, None, 6 8256 activation_34[0][0]
__________________________________________________________________________________________________
batch_normalization_35 (BatchNo (None, None, None, 6 256 conv2d_35[0][0]
__________________________________________________________________________________________________
activation_35 (Activation) (None, None, None, 6 0 batch_normalization_35[0][0]
__________________________________________________________________________________________________
conv2d_36 (Conv2D) (None, None, None, 1 25616 activation_35[0][0]
__________________________________________________________________________________________________
concatenate_16 (Concatenate) (None, None, None, 2 0 max_pooling2d_4[0][0]
concatenate_13[0][0]
concatenate_14[0][0]
concatenate_15[0][0]
conv2d_36[0][0]
__________________________________________________________________________________________________
batch_normalization_36 (BatchNo (None, None, None, 2 1024 concatenate_16[0][0]
__________________________________________________________________________________________________
activation_36 (Activation) (None, None, None, 2 0 batch_normalization_36[0][0]
__________________________________________________________________________________________________
conv2d_37 (Conv2D) (None, None, None, 1 4112 activation_36[0][0]
__________________________________________________________________________________________________
up_sampling2d_1 (UpSampling2D) (None, None, None, 1 0 conv2d_37[0][0]
__________________________________________________________________________________________________
concatenate_17 (Concatenate) (None, None, None, 3 0 up_sampling2d_1[0][0]
conv2d_28[0][0]
__________________________________________________________________________________________________
conv2d_transpose_1 (Conv2DTrans (None, None, None, 3 25632 concatenate_17[0][0]
__________________________________________________________________________________________________
activation_37 (Activation) (None, None, None, 3 0 conv2d_transpose_1[0][0]
__________________________________________________________________________________________________
batch_normalization_37 (BatchNo (None, None, None, 3 128 activation_37[0][0]
__________________________________________________________________________________________________
conv2d_transpose_2 (Conv2DTrans (None, None, None, 1 528 batch_normalization_37[0][0]
__________________________________________________________________________________________________
activation_38 (Activation) (None, None, None, 1 0 conv2d_transpose_2[0][0]
__________________________________________________________________________________________________
batch_normalization_38 (BatchNo (None, None, None, 1 64 activation_38[0][0]
__________________________________________________________________________________________________
up_sampling2d_2 (UpSampling2D) (None, None, None, 1 0 batch_normalization_38[0][0]
__________________________________________________________________________________________________
concatenate_18 (Concatenate) (None, None, None, 3 0 up_sampling2d_2[0][0]
conv2d_19[0][0]
__________________________________________________________________________________________________
conv2d_transpose_3 (Conv2DTrans (None, None, None, 3 25632 concatenate_18[0][0]
__________________________________________________________________________________________________
activation_39 (Activation) (None, None, None, 3 0 conv2d_transpose_3[0][0]
__________________________________________________________________________________________________
batch_normalization_39 (BatchNo (None, None, None, 3 128 activation_39[0][0]
__________________________________________________________________________________________________
conv2d_transpose_4 (Conv2DTrans (None, None, None, 1 528 batch_normalization_39[0][0]
__________________________________________________________________________________________________
activation_40 (Activation) (None, None, None, 1 0 conv2d_transpose_4[0][0]
__________________________________________________________________________________________________
batch_normalization_40 (BatchNo (None, None, None, 1 64 activation_40[0][0]
__________________________________________________________________________________________________
up_sampling2d_3 (UpSampling2D) (None, None, None, 1 0 batch_normalization_40[0][0]
__________________________________________________________________________________________________
concatenate_19 (Concatenate) (None, None, None, 3 0 up_sampling2d_3[0][0]
conv2d_10[0][0]
__________________________________________________________________________________________________
conv2d_transpose_5 (Conv2DTrans (None, None, None, 3 25632 concatenate_19[0][0]
__________________________________________________________________________________________________
activation_41 (Activation) (None, None, None, 3 0 conv2d_transpose_5[0][0]
__________________________________________________________________________________________________
batch_normalization_41 (BatchNo (None, None, None, 3 128 activation_41[0][0]
__________________________________________________________________________________________________
conv2d_transpose_6 (Conv2DTrans (None, None, None, 1 528 batch_normalization_41[0][0]
__________________________________________________________________________________________________
activation_42 (Activation) (None, None, None, 1 0 conv2d_transpose_6[0][0]
__________________________________________________________________________________________________
batch_normalization_42 (BatchNo (None, None, None, 1 64 activation_42[0][0]
__________________________________________________________________________________________________
up_sampling2d_4 (UpSampling2D) (None, None, None, 1 0 batch_normalization_42[0][0]
__________________________________________________________________________________________________
concatenate_20 (Concatenate) (None, None, None, 3 0 up_sampling2d_4[0][0]
conv2d_1[0][0]
__________________________________________________________________________________________________
conv2d_transpose_7 (Conv2DTrans (None, None, None, 3 25632 concatenate_20[0][0]
__________________________________________________________________________________________________
activation_43 (Activation) (None, None, None, 3 0 conv2d_transpose_7[0][0]
__________________________________________________________________________________________________
conv2d_transpose_8 (Conv2DTrans (None, None, None, 1 33 activation_43[0][0]
__________________________________________________________________________________________________
subtract_1 (Subtract) (None, None, None, 1 0 input_1[0][0]
conv2d_transpose_8[0][0]
==================================================================================================
Total params: 606,321
Trainable params: 600,017
Non-trainable params: 6,304
__________________________________________________________________________________________________
history=ddnet_model.fit(sparseview_60_train, ndct_train, validation_split=0.1, batch_size=10, epochs=100)
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:1033: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.
WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:1020: The name tf.assign is deprecated. Please use tf.compat.v1.assign instead.
Train on 540 samples, validate on 60 samples
Epoch 1/100
540/540 [==============================] - 43s 80ms/step - loss: 0.0280 - tf_psnr: 21.2017 - val_loss: 0.0075 - val_tf_psnr: 21.2391
Epoch 2/100
540/540 [==============================] - 26s 49ms/step - loss: 0.0019 - tf_psnr: 27.2900 - val_loss: 0.0018 - val_tf_psnr: 27.3542
Epoch 3/100
540/540 [==============================] - 26s 49ms/step - loss: 0.0014 - tf_psnr: 28.5661 - val_loss: 0.0013 - val_tf_psnr: 28.8296
Epoch 4/100
540/540 [==============================] - 26s 49ms/step - loss: 0.0011 - tf_psnr: 29.5215 - val_loss: 0.0013 - val_tf_psnr: 29.0090
Epoch 5/100
540/540 [==============================] - 26s 49ms/step - loss: 0.0010 - tf_psnr: 29.9749 - val_loss: 0.0010 - val_tf_psnr: 29.8063
Epoch 6/100
540/540 [==============================] - 26s 49ms/step - loss: 0.0011 - tf_psnr: 29.6950 - val_loss: 8.8809e-04 - val_tf_psnr: 30.5203
Epoch 7/100
540/540 [==============================] - 26s 49ms/step - loss: 0.0011 - tf_psnr: 29.8324 - val_loss: 0.0018 - val_tf_psnr: 27.3881
Epoch 8/100
540/540 [==============================] - 26s 49ms/step - loss: 9.7362e-04 - tf_psnr: 30.1941 - val_loss: 7.9895e-04 - val_tf_psnr: 30.9795
Epoch 9/100
540/540 [==============================] - 26s 49ms/step - loss: 8.2528e-04 - tf_psnr: 30.8696 - val_loss: 7.3486e-04 - val_tf_psnr: 31.3439
Epoch 10/100
540/540 [==============================] - 26s 49ms/step - loss: 7.5115e-04 - tf_psnr: 31.2677 - val_loss: 7.1803e-04 - val_tf_psnr: 31.4459
Epoch 11/100
540/540 [==============================] - 26s 49ms/step - loss: 8.1593e-04 - tf_psnr: 30.9633 - val_loss: 7.8475e-04 - val_tf_psnr: 31.0650
Epoch 12/100
540/540 [==============================] - 26s 49ms/step - loss: 7.6514e-04 - tf_psnr: 31.2249 - val_loss: 5.9724e-04 - val_tf_psnr: 32.2474
Epoch 13/100
540/540 [==============================] - 26s 49ms/step - loss: 6.6273e-04 - tf_psnr: 31.8184 - val_loss: 5.6468e-04 - val_tf_psnr: 32.4912
Epoch 14/100
540/540 [==============================] - 26s 49ms/step - loss: 7.3347e-04 - tf_psnr: 31.4385 - val_loss: 7.6936e-04 - val_tf_psnr: 31.1439
Epoch 15/100
540/540 [==============================] - 26s 49ms/step - loss: 6.9742e-04 - tf_psnr: 31.6504 - val_loss: 5.4058e-04 - val_tf_psnr: 32.6876
Epoch 16/100
540/540 [==============================] - 26s 49ms/step - loss: 8.2124e-04 - tf_psnr: 31.0641 - val_loss: 5.3755e-04 - val_tf_psnr: 32.7044
Epoch 17/100
540/540 [==============================] - 26s 49ms/step - loss: 7.1717e-04 - tf_psnr: 31.5721 - val_loss: 5.0459e-04 - val_tf_psnr: 32.9812
Epoch 18/100
540/540 [==============================] - 26s 49ms/step - loss: 5.4516e-04 - tf_psnr: 32.6510 - val_loss: 4.7683e-04 - val_tf_psnr: 33.2336
Epoch 19/100
540/540 [==============================] - 26s 49ms/step - loss: 5.5237e-04 - tf_psnr: 32.5945 - val_loss: 4.8046e-04 - val_tf_psnr: 33.1914
Epoch 20/100
540/540 [==============================] - 26s 49ms/step - loss: 5.7488e-04 - tf_psnr: 32.4477 - val_loss: 7.0300e-04 - val_tf_psnr: 31.5445
Epoch 21/100
540/540 [==============================] - 26s 49ms/step - loss: 5.9892e-04 - tf_psnr: 32.3263 - val_loss: 4.8432e-04 - val_tf_psnr: 33.1533
Epoch 22/100
540/540 [==============================] - 26s 49ms/step - loss: 5.6773e-04 - tf_psnr: 32.5223 - val_loss: 4.0766e-04 - val_tf_psnr: 33.9022
Epoch 23/100
540/540 [==============================] - 26s 49ms/step - loss: 5.7343e-04 - tf_psnr: 32.4803 - val_loss: 4.9867e-04 - val_tf_psnr: 33.0359
Epoch 24/100
540/540 [==============================] - 26s 49ms/step - loss: 5.2029e-04 - tf_psnr: 32.8887 - val_loss: 4.1754e-04 - val_tf_psnr: 33.7978
Epoch 25/100
540/540 [==============================] - 26s 49ms/step - loss: 5.1400e-04 - tf_psnr: 32.9858 - val_loss: 4.8430e-04 - val_tf_psnr: 33.1594
Epoch 26/100
540/540 [==============================] - 26s 49ms/step - loss: 4.9536e-04 - tf_psnr: 33.1164 - val_loss: 4.1719e-04 - val_tf_psnr: 33.7993
Epoch 27/100
540/540 [==============================] - 26s 49ms/step - loss: 4.8681e-04 - tf_psnr: 33.1766 - val_loss: 3.8200e-04 - val_tf_psnr: 34.1862
Epoch 28/100
540/540 [==============================] - 26s 49ms/step - loss: 4.5471e-04 - tf_psnr: 33.4699 - val_loss: 4.0245e-04 - val_tf_psnr: 33.9595
Epoch 29/100
540/540 [==============================] - 26s 49ms/step - loss: 4.7532e-04 - tf_psnr: 33.3320 - val_loss: 3.6850e-04 - val_tf_psnr: 34.3388
Epoch 30/100
540/540 [==============================] - 26s 49ms/step - loss: 4.2976e-04 - tf_psnr: 33.7415 - val_loss: 3.5544e-04 - val_tf_psnr: 34.5017
Epoch 31/100
540/540 [==============================] - 26s 49ms/step - loss: 3.7445e-04 - tf_psnr: 34.2973 - val_loss: 3.5161e-04 - val_tf_psnr: 34.5458
Epoch 32/100
540/540 [==============================] - 26s 49ms/step - loss: 4.1256e-04 - tf_psnr: 33.9205 - val_loss: 3.4630e-04 - val_tf_psnr: 34.6239
Epoch 33/100
540/540 [==============================] - 26s 49ms/step - loss: 3.8373e-04 - tf_psnr: 34.2091 - val_loss: 3.3894e-04 - val_tf_psnr: 34.7115
Epoch 34/100
540/540 [==============================] - 26s 49ms/step - loss: 3.7360e-04 - tf_psnr: 34.3421 - val_loss: 4.5653e-04 - val_tf_psnr: 33.4073
Epoch 35/100
540/540 [==============================] - 26s 48ms/step - loss: 3.8419e-04 - tf_psnr: 34.2131 - val_loss: 4.2897e-04 - val_tf_psnr: 33.6842
Epoch 36/100
540/540 [==============================] - 26s 49ms/step - loss: 4.0863e-04 - tf_psnr: 33.9573 - val_loss: 4.4508e-04 - val_tf_psnr: 33.5211
Epoch 37/100
540/540 [==============================] - 26s 49ms/step - loss: 3.9134e-04 - tf_psnr: 34.1043 - val_loss: 3.7600e-04 - val_tf_psnr: 34.2608
Epoch 38/100
540/540 [==============================] - 26s 49ms/step - loss: 4.2303e-04 - tf_psnr: 33.8016 - val_loss: 4.1222e-04 - val_tf_psnr: 33.8547
Epoch 39/100
540/540 [==============================] - 26s 49ms/step - loss: 3.7695e-04 - tf_psnr: 34.3062 - val_loss: 3.5216e-04 - val_tf_psnr: 34.5391
Epoch 40/100
540/540 [==============================] - 26s 49ms/step - loss: 3.4057e-04 - tf_psnr: 34.7122 - val_loss: 3.1190e-04 - val_tf_psnr: 35.0734
Epoch 41/100
540/540 [==============================] - 26s 49ms/step - loss: 3.6780e-04 - tf_psnr: 34.3927 - val_loss: 4.7632e-04 - val_tf_psnr: 33.2255
Epoch 42/100
540/540 [==============================] - 26s 49ms/step - loss: 3.7832e-04 - tf_psnr: 34.2659 - val_loss: 3.3005e-04 - val_tf_psnr: 34.8223
Epoch 43/100
540/540 [==============================] - 26s 49ms/step - loss: 4.8456e-04 - tf_psnr: 33.5204 - val_loss: 0.0012 - val_tf_psnr: 29.3339
Epoch 44/100
540/540 [==============================] - 26s 49ms/step - loss: 4.1001e-04 - tf_psnr: 34.1469 - val_loss: 3.0927e-04 - val_tf_psnr: 35.1152
Epoch 45/100
540/540 [==============================] - 26s 49ms/step - loss: 3.5253e-04 - tf_psnr: 34.6016 - val_loss: 3.7163e-04 - val_tf_psnr: 34.3030
Epoch 46/100
540/540 [==============================] - 26s 49ms/step - loss: 3.7146e-04 - tf_psnr: 34.4000 - val_loss: 4.9450e-04 - val_tf_psnr: 33.0694
Epoch 47/100
540/540 [==============================] - 26s 49ms/step - loss: 3.8757e-04 - tf_psnr: 34.2648 - val_loss: 3.0528e-04 - val_tf_psnr: 35.1773
Epoch 48/100
540/540 [==============================] - 26s 49ms/step - loss: 2.8734e-04 - tf_psnr: 35.4420 - val_loss: 2.7628e-04 - val_tf_psnr: 35.6151
Epoch 49/100
540/540 [==============================] - 26s 49ms/step - loss: 3.0102e-04 - tf_psnr: 35.2569 - val_loss: 2.6695e-04 - val_tf_psnr: 35.7558
Epoch 50/100
540/540 [==============================] - 26s 49ms/step - loss: 2.9944e-04 - tf_psnr: 35.2682 - val_loss: 3.3417e-04 - val_tf_psnr: 34.7941
Epoch 51/100
540/540 [==============================] - 26s 49ms/step - loss: 3.3042e-04 - tf_psnr: 34.8676 - val_loss: 3.6515e-04 - val_tf_psnr: 34.3759
Epoch 52/100
540/540 [==============================] - 26s 49ms/step - loss: 3.2407e-04 - tf_psnr: 34.9452 - val_loss: 3.1161e-04 - val_tf_psnr: 35.0818
Epoch 53/100
540/540 [==============================] - 26s 49ms/step - loss: 3.4927e-04 - tf_psnr: 34.6836 - val_loss: 3.1465e-04 - val_tf_psnr: 35.0449
Epoch 54/100
540/540 [==============================] - 26s 49ms/step - loss: 2.8681e-04 - tf_psnr: 35.4498 - val_loss: 3.1488e-04 - val_tf_psnr: 35.0326
Epoch 55/100
540/540 [==============================] - 26s 49ms/step - loss: 2.9092e-04 - tf_psnr: 35.3946 - val_loss: 3.0132e-04 - val_tf_psnr: 35.2409
Epoch 56/100
540/540 [==============================] - 26s 49ms/step - loss: 2.6628e-04 - tf_psnr: 35.7641 - val_loss: 3.1757e-04 - val_tf_psnr: 35.0018
Epoch 57/100
540/540 [==============================] - 26s 49ms/step - loss: 3.4802e-04 - tf_psnr: 34.6686 - val_loss: 3.3183e-04 - val_tf_psnr: 34.7981
Epoch 58/100
540/540 [==============================] - 26s 49ms/step - loss: 2.9418e-04 - tf_psnr: 35.3574 - val_loss: 2.8899e-04 - val_tf_psnr: 35.4286
Epoch 59/100
540/540 [==============================] - 26s 49ms/step - loss: 2.8491e-04 - tf_psnr: 35.5120 - val_loss: 3.0041e-04 - val_tf_psnr: 35.2706
Epoch 60/100
540/540 [==============================] - 26s 49ms/step - loss: 2.7682e-04 - tf_psnr: 35.6094 - val_loss: 2.6075e-04 - val_tf_psnr: 35.8742
Epoch 61/100
540/540 [==============================] - 26s 49ms/step - loss: 2.5840e-04 - tf_psnr: 35.8941 - val_loss: 2.5724e-04 - val_tf_psnr: 35.9425
Epoch 62/100
540/540 [==============================] - 26s 49ms/step - loss: 2.8665e-04 - tf_psnr: 35.4623 - val_loss: 2.9329e-04 - val_tf_psnr: 35.3503
Epoch 63/100
540/540 [==============================] - 26s 49ms/step - loss: 2.8515e-04 - tf_psnr: 35.4931 - val_loss: 3.0674e-04 - val_tf_psnr: 35.1669
Epoch 64/100
540/540 [==============================] - 26s 49ms/step - loss: 3.1575e-04 - tf_psnr: 35.0711 - val_loss: 2.6852e-04 - val_tf_psnr: 35.7359
Epoch 65/100
540/540 [==============================] - 26s 49ms/step - loss: 2.6189e-04 - tf_psnr: 35.8535 - val_loss: 2.8497e-04 - val_tf_psnr: 35.4618
Epoch 66/100
540/540 [==============================] - 26s 49ms/step - loss: 2.6955e-04 - tf_psnr: 35.7327 - val_loss: 3.1691e-04 - val_tf_psnr: 35.0373
Epoch 67/100
540/540 [==============================] - 26s 48ms/step - loss: 2.8370e-04 - tf_psnr: 35.5343 - val_loss: 2.4236e-04 - val_tf_psnr: 36.2066
Epoch 68/100
540/540 [==============================] - 26s 49ms/step - loss: 2.5478e-04 - tf_psnr: 35.9745 - val_loss: 2.4796e-04 - val_tf_psnr: 36.1104
Epoch 69/100
540/540 [==============================] - 26s 48ms/step - loss: 2.8271e-04 - tf_psnr: 35.5821 - val_loss: 2.9078e-04 - val_tf_psnr: 35.3868
Epoch 70/100
540/540 [==============================] - 26s 48ms/step - loss: 3.2817e-04 - tf_psnr: 34.9521 - val_loss: 2.6680e-04 - val_tf_psnr: 35.7780
Epoch 71/100
540/540 [==============================] - 26s 48ms/step - loss: 2.8813e-04 - tf_psnr: 35.4786 - val_loss: 3.0792e-04 - val_tf_psnr: 35.1429
Epoch 72/100
540/540 [==============================] - 26s 48ms/step - loss: 2.5403e-04 - tf_psnr: 35.9827 - val_loss: 2.5393e-04 - val_tf_psnr: 36.0012
Epoch 73/100
540/540 [==============================] - 26s 48ms/step - loss: 2.4149e-04 - tf_psnr: 36.1899 - val_loss: 2.4901e-04 - val_tf_psnr: 36.0830
Epoch 74/100
540/540 [==============================] - 26s 48ms/step - loss: 2.3129e-04 - tf_psnr: 36.3773 - val_loss: 2.4793e-04 - val_tf_psnr: 36.1201
Epoch 75/100
540/540 [==============================] - 26s 48ms/step - loss: 2.9249e-04 - tf_psnr: 35.5059 - val_loss: 3.3721e-04 - val_tf_psnr: 34.7488
Epoch 76/100
540/540 [==============================] - 26s 48ms/step - loss: 2.6651e-04 - tf_psnr: 35.7920 - val_loss: 2.5249e-04 - val_tf_psnr: 36.0162
Epoch 77/100
540/540 [==============================] - 26s 48ms/step - loss: 2.6189e-04 - tf_psnr: 35.8730 - val_loss: 3.0696e-04 - val_tf_psnr: 35.1525
Epoch 78/100
540/540 [==============================] - 26s 48ms/step - loss: 2.3019e-04 - tf_psnr: 36.4130 - val_loss: 2.5425e-04 - val_tf_psnr: 35.9807
Epoch 79/100
540/540 [==============================] - 26s 48ms/step - loss: 2.2941e-04 - tf_psnr: 36.4208 - val_loss: 2.5265e-04 - val_tf_psnr: 36.0124
Epoch 80/100
540/540 [==============================] - 26s 49ms/step - loss: 2.2735e-04 - tf_psnr: 36.4574 - val_loss: 2.6618e-04 - val_tf_psnr: 35.7907
Epoch 81/100
540/540 [==============================] - 26s 48ms/step - loss: 2.5631e-04 - tf_psnr: 35.9504 - val_loss: 2.4669e-04 - val_tf_psnr: 36.1197
Epoch 82/100
540/540 [==============================] - 26s 48ms/step - loss: 2.3399e-04 - tf_psnr: 36.3353 - val_loss: 2.3497e-04 - val_tf_psnr: 36.3390
Epoch 83/100
540/540 [==============================] - 26s 48ms/step - loss: 2.4020e-04 - tf_psnr: 36.2282 - val_loss: 3.6146e-04 - val_tf_psnr: 34.4420
Epoch 84/100
540/540 [==============================] - 26s 48ms/step - loss: 2.4922e-04 - tf_psnr: 36.0688 - val_loss: 2.5018e-04 - val_tf_psnr: 36.0507
Epoch 85/100
540/540 [==============================] - 26s 48ms/step - loss: 2.2389e-04 - tf_psnr: 36.5251 - val_loss: 2.4655e-04 - val_tf_psnr: 36.1217
Epoch 86/100
540/540 [==============================] - 26s 48ms/step - loss: 2.3999e-04 - tf_psnr: 36.2380 - val_loss: 2.4482e-04 - val_tf_psnr: 36.1537
Epoch 87/100
540/540 [==============================] - 26s 48ms/step - loss: 2.5460e-04 - tf_psnr: 35.9957 - val_loss: 2.6136e-04 - val_tf_psnr: 35.8773
Epoch 88/100
540/540 [==============================] - 26s 48ms/step - loss: 2.1452e-04 - tf_psnr: 36.7124 - val_loss: 2.3646e-04 - val_tf_psnr: 36.3289
Epoch 89/100
540/540 [==============================] - 26s 48ms/step - loss: 2.0984e-04 - tf_psnr: 36.7986 - val_loss: 2.4461e-04 - val_tf_psnr: 36.1807
Epoch 90/100
540/540 [==============================] - 26s 48ms/step - loss: 2.1961e-04 - tf_psnr: 36.6152 - val_loss: 2.3673e-04 - val_tf_psnr: 36.3248
Epoch 91/100
540/540 [==============================] - 26s 48ms/step - loss: 2.1523e-04 - tf_psnr: 36.6929 - val_loss: 2.4572e-04 - val_tf_psnr: 36.1452
Epoch 92/100
540/540 [==============================] - 26s 48ms/step - loss: 2.8878e-04 - tf_psnr: 35.5544 - val_loss: 2.9032e-04 - val_tf_psnr: 35.3993
Epoch 93/100
540/540 [==============================] - 26s 48ms/step - loss: 3.0156e-04 - tf_psnr: 35.2996 - val_loss: 3.7494e-04 - val_tf_psnr: 34.2842
Epoch 94/100
540/540 [==============================] - 26s 48ms/step - loss: 2.2183e-04 - tf_psnr: 36.5679 - val_loss: 2.4650e-04 - val_tf_psnr: 36.1272
Epoch 95/100
540/540 [==============================] - 26s 48ms/step - loss: 2.0715e-04 - tf_psnr: 36.8539 - val_loss: 2.2339e-04 - val_tf_psnr: 36.5729
Epoch 96/100
540/540 [==============================] - 26s 48ms/step - loss: 2.1323e-04 - tf_psnr: 36.7401 - val_loss: 2.1736e-04 - val_tf_psnr: 36.7067
Epoch 97/100
540/540 [==============================] - 26s 48ms/step - loss: 2.0388e-04 - tf_psnr: 36.9348 - val_loss: 2.3284e-04 - val_tf_psnr: 36.3830
Epoch 98/100
540/540 [==============================] - 26s 48ms/step - loss: 2.0425e-04 - tf_psnr: 36.9245 - val_loss: 2.4491e-04 - val_tf_psnr: 36.1792
Epoch 99/100
540/540 [==============================] - 26s 48ms/step - loss: 2.1531e-04 - tf_psnr: 36.6909 - val_loss: 2.4497e-04 - val_tf_psnr: 36.1846
Epoch 100/100
540/540 [==============================] - 26s 48ms/step - loss: 2.1190e-04 - tf_psnr: 36.7692 - val_loss: 2.4295e-04 - val_tf_psnr: 36.2184
Save weights for the future reuse.
ddnet_model.save_weights("/content/drive/My Drive/Colab Notebooks/dd_net/model.h5")
Plotting PSNR values to see the trend.
import matplotlib.pyplot as plt
acc = history.history['tf_psnr']
val_acc = history.history['val_tf_psnr']
loss = history.history['loss']
val_loss = history.history['val_loss']
epochs = range(len(acc))
plt.plot(epochs, acc, 'r', label='Training accuracy')
plt.plot(epochs, val_acc, 'b', label='Validation accuracy')
plt.title('Training and validation accuracy')
plt.legend(loc=0)
plt.figure()
plt.show()
<Figure size 432x288 with 0 Axes>
reconstructed = ddnet_model.predict(sparseview_60_test)
psnr = cal_psnr(ndct_test, reconstructed)
print("psnr 40 epochs.....",psnr)
psnr 40 epochs..... 33.99149254705274
psnr 40 epochs….. 33.3782823500309
from PIL import Image
a = reconstructed[0].reshape(512, 512)
scalef = np.amax(a)
a = np.clip(255 * a/scalef, 0, 255).astype('uint8')
#result = Image.fromarray((a * 255).astype(np.uint8))
result = Image.fromarray((a).astype(np.uint8))
# result.save('unet_15_600_0.png')
result.save('/content/drive/My Drive/Colab Notebooks/dd_net/reconstructed_ddnet_0.png')
a = reconstructed[99].reshape(512, 512)
scalef = np.amax(a)
a = np.clip(255 * a/scalef, 0, 255).astype('uint8')
#result = Image.fromarray((a * 255).astype(np.uint8))
result = Image.fromarray((a).astype(np.uint8))
# result.save('unet_15_600_99.png')
result.save('/content/drive/My Drive/Colab Notebooks/dd_net/reconstructed_ddnet_99.png')
a = sparseview_60_test[0].reshape(512, 512)
scalef = np.amax(a)
a = np.clip(255 * a/scalef, 0, 255).astype('uint8')
#result = Image.fromarray((a * 255).astype(np.uint8))
result = Image.fromarray((a).astype(np.uint8))
# result.save('unet_15_600_0.png')
result.save('/content/drive/My Drive/Colab Notebooks/dd_net/sparseview_60_test_ddnet_0.png')
a = sparseview_60_test[99].reshape(512, 512)
scalef = np.amax(a)
a = np.clip(255 * a/scalef, 0, 255).astype('uint8')
#result = Image.fromarray((a * 255).astype(np.uint8))
result = Image.fromarray((a).astype(np.uint8))
# result.save('unet_15_600_99.png')
result.save('/content/drive/My Drive/Colab Notebooks/dd_net/sparseview_60_test_ddnet_99.png')
# Plot training & validation accuracy values
plt.plot(history.history['tf_psnr'])
plt.plot(history.history['val_tf_psnr'])
plt.title('Model accuracy')
plt.ylabel('Accuracy')
plt.xlabel('Epoch')
plt.legend(['Train', 'Test'], loc='upper left')
plt.show()
# Plot training & validation loss values
plt.plot(history.history['loss'])
plt.plot(history.history['val_loss'])
plt.title('Model loss')
plt.ylabel('Loss')
plt.xlabel('Epoch')
plt.legend(['Train', 'Test'], loc='upper left')
plt.show()
img=mpimg.imread('/content/drive/My Drive/Colab Notebooks/dd_net/sparseview_60_test_ddnet_0.png')
imgplot = plt.imshow(img)
plt.show()
img=mpimg.imread('/content/drive/My Drive/Colab Notebooks/dd_net/reconstructed_ddnet_0.png')
imgplot = plt.imshow(img)
plt.show()
img=mpimg.imread('/content/drive/My Drive/Colab Notebooks/dd_net/sparseview_60_test_ddnet_99.png')
imgplot = plt.imshow(img)
plt.show()
img=mpimg.imread('/content/drive/My Drive/Colab Notebooks/dd_net/reconstructed_ddnet_99.png')
imgplot = plt.imshow(img)
plt.show()
Images
sparseview_60_0
Reconstructed sparseview_60_0
sparseview_60_99
Reconstructed sparseview_60_99
