how do I scatter a tensor with smaller shape to the bigger tensor?

Question

I have an input tensor with shape (1,4,4,1), I want to concatenate this input with the output of conv layer with shape (1,28,28,1), but for this I want to put each value of the input to the new tensor with shape (1,28,28,1) so that this value put in the middle of 7x7 blocks. if I had an np array I do this with: w_expand[:, 3::7, 3::7] = wt_random but with tensor and lambda layer in keras, I do not know what should I do? could you please help me with this issue?

from keras.layers import Input, Concatenate, GaussianNoise,Dropout,BatchNormalization,MaxPool2D,AveragePooling2D
from keras.layers import Conv2D, AtrousConv2D
from keras.models import Model
from keras.datasets import mnist
from keras.callbacks import TensorBoard
from keras import backend as K
from keras import layers
import matplotlib.pyplot as plt
import tensorflow as tf
import keras as Kr
from keras.optimizers import SGD,RMSprop,Adam
from keras.callbacks import ReduceLROnPlateau
from keras.callbacks import EarlyStopping
from keras.callbacks import ModelCheckpoint
import numpy as np
import pylab as pl
import matplotlib.cm as cm
import keract
from matplotlib import pyplot
from keras import optimizers
from keras import regularizers

from tensorflow.python.keras.layers import Lambda;
#-----------------building w train---------------------------------------------
wt_random=np.random.randint(2, size=(49999,4,4))
w_expand=wt_random.astype(np.float32)
wv_random=np.random.randint(2, size=(9999,4,4))
wv_expand=wv_random.astype(np.float32)
x,y,z=w_expand.shape
w_expand=w_expand.reshape((x,y,z,1))
x,y,z=wv_expand.shape
wv_expand=wv_expand.reshape((x,y,z,1))

#-----------------building w test---------------------------------------------
w_test = np.random.randint(2,size=(1,4,4))
w_test=w_test.astype(np.float32)
w_test=w_test.reshape((1,4,4,1))


#-----------------------encoder------------------------------------------------
#------------------------------------------------------------------------------
wtm=Input((4,4,1))
image = Input((28, 28, 1))
conv1 = Conv2D(64, (5, 5), activation='relu', padding='same', name='convl1e',dilation_rate=(2,2))(image)
conv2 = Conv2D(64, (5, 5), activation='relu', padding='same', name='convl2e',dilation_rate=(2,2))(conv1)
conv3 = Conv2D(64, (5, 5), activation='relu', padding='same', name='convl3e',dilation_rate=(2,2))(conv2)
BN=BatchNormalization()(conv3)
encoded =  Conv2D(1, (5, 5), activation='relu', padding='same',name='encoded_I')(BN)

#-----------------------adding w---------------------------------------

I do not know how do I make wtmp with lambda layer in this part????
encoded_merged = Concatenate(axis=3)([encoded, wtmp])

#-----------------------decoder------------------------------------------------
#------------------------------------------------------------------------------
#deconv_input=Input((28,28,1),name='inputTodeconv')
#encoded_merged = Input((28, 28, 2))
deconv1 = Conv2D(64, (5, 5), activation='relu', padding='same', name='convl1d',dilation_rate=(2,2))(encoded_merged)
deconv2 = Conv2D(64, (5, 5), activation='relu', padding='same', name='convl2d',dilation_rate=(2,2))(deconv1)
deconv3 = Conv2D(64, (5, 5), activation='relu',padding='same', name='convl3d',dilation_rate=(2,2))(deconv2)
deconv4 = Conv2D(64, (5, 5), activation='relu',padding='same', name='convl4d',dilation_rate=(2,2))(deconv3)
BNd=BatchNormalization()(deconv4)

decoded = Conv2D(1, (5, 5), activation='sigmoid', padding='same', name='decoder_output')(BNd) 

model=Model(inputs=[image,wtm],outputs=decoded)

decoded_noise = GaussianNoise(0.5)(decoded)

#----------------------w extraction------------------------------------
convw1 = Conv2D(64, (5,5), activation='relu', name='conl1w')(decoded_noise)#24
convw2 = Conv2D(64, (5,5), activation='relu', name='convl2w')(convw1)#20
convw3 = Conv2D(64, (5,5), activation='relu' ,name='conl3w')(convw2)#16
convw4 = Conv2D(64, (5,5), activation='relu' ,name='conl4w')(convw3)#12
convw5 = Conv2D(64, (5,5), activation='relu', name='conl5w')(convw4)#8
convw6 = Conv2D(64, (5,5), activation='relu', name='conl6w')(convw5)#4
BNed=BatchNormalization()(convw6)
pred_w = Conv2D(1, (1, 1), activation='sigmoid', padding='same', name='reconstructed_W',dilation_rate=(2,2))(BNed)  
w=Model(inputs=[image,wtm],outputs=[decoded,pred_w])

w.summary()

This seems to be a duplicate of: https://stackoverflow.com/questions/54918798/tile-a-tensor-from-a-smaller-shape-to-a-bigger-one/54920998#54920998 — tomkot, Mar 06 '19 at 07:39
thank you, but they are different. here I need to first consider a tensor with shape (1,28,28,1) instead (1,4,4,1) and then put each value of smaller tensor in the middle pixel of 7x7 block of the bigger tensor. but in the mentioned link the (1,4,41) tensor is repeated to make the (1,28,28,1) tensor. — david, Mar 06 '19 at 15:14
could you please help me with this problem? as I said before if it was a numpy array I can do this with w_expand[:, 3::7, 3::7] = wt_random but with tensor and lambda layer I do not know what should I do? — david, Mar 06 '19 at 15:14

how do I scatter a tensor with smaller shape to the bigger tensor?

0 Answers0