In Theano, how to do Reverse-MaxPooling in the Convolutional MaxPooling Auto-Encoder

MaxPooling Layer always follows Convolutional Layer. But in the Convolutional MaxPooling Auto-Encoder, how to do reverse or Un-MaxPooling?

Here is the method:

import numpy as np
import theano
import theano.tensor as T
import theano.sandbox.neighbours as nbr

poolsize = (2, 2)
print 'neighbor box:'
print poolsize
# Defining batch images
images = T.tensor4('images')
#conver images to neighbors matrix
neibs = nbr.images2neibs(images, neib_shape=poolsize)

# Constructing theano function 
window_function = theano.function([images], neibs)

# Input tensor image width and height are 4
im_val = np.random.randint(0, 100, (2, 1, 4, 4))
im_val = im_val.astype(np.float32)
# Function application
neibs_val = window_function(im_val)

print 'original images:'
print im_val
print 'neighbor flatten:'
print neibs_val


neibs0 = T.zeros_like(neibs)
#get maximum of neighbors
am = T.argmax(neibs, axis=1)
neibs1 = T.set_subtensor(neibs0[T.arange(neibs.shape[0]), am], 1)
neibs *= neibs1 #suppress the non-maximum

#convert neighbors matrix to images
im_new = nbr.neibs2images(neibs, poolsize, im_val.shape)
# Theano function definition
inv_window = theano.function([images], im_new)
# Function application
im_new_val = inv_window(im_val)

print 'non maximum suppression result:'
print im_new_val

The following is the output:

neighbor box:
(2, 2)
original images:
[[[[ 15.  46.  46.  37.]
   [ 73.  93.   2.  91.]
   [ 97.  93.  78.   7.]
   [  0.  24.  34.  26.]]]


 [[[ 70.  46.   2.  93.]
   [ 11.  54.  29.  83.]
   [  0.  48.  69.  95.]
   [ 49.  48.  10.  19.]]]]
neighbor flatten:
[[ 15.  46.  73.  93.]
 [ 46.  37.   2.  91.]
 [ 97.  93.   0.  24.]
 [ 78.   7.  34.  26.]
 [ 70.  46.  11.  54.]
 [  2.  93.  29.  83.]
 [  0.  48.  49.  48.]
 [ 69.  95.  10.  19.]]
non maximum suppression result:
[[[[  0.   0.   0.   0.]
   [  0.  93.   0.  91.]
   [ 97.   0.  78.   0.]
   [  0.   0.   0.   0.]]]


 [[[ 70.   0.   0.  93.]
   [  0.   0.   0.   0.]
   [  0.   0.   0.  95.]
   [ 49.   0.   0.   0.]]]]