sandbox.neighbours – Neighbours Ops

API

Neighbours was moved into theano.tensor.nnet.neighbours. This file was created for compatibility.

theano.sandbox.neighbours.images2neibs(ten4, neib_shape, neib_step=None, mode='valid')

Function images2neibs allows to apply a sliding window operation to a tensor containing images or other two-dimensional objects. The sliding window operation loops over points in input data and stores a rectangular neighbourhood of each point. It is possible to assign a step of selecting patches (parameter neib_step).

Parameters:
  • ten4 (A 4d tensor-like) – A 4-dimensional tensor which represents a list of lists of images. It should have shape (list 1 dim, list 2 dim, row, col). The first two dimensions can be useful to store different channels and batches.
  • neib_shape (A 1d tensor-like of 2 values) – A tuple containing two values: height and width of the neighbourhood. It should have shape (r,c) where r is the height of the neighborhood in rows and c is the width of the neighborhood in columns.
  • neib_step (A 1d tensor-like of 2 values) –

    (dr,dc) where dr is the number of rows to skip between patch and dc is the number of columns. The parameter should be a tuple of two elements: number of rows and number of columns to skip each iteration. Basically, when the step is 1, the neighbourhood of every first element is taken and every possible rectangular subset is returned. By default it is equal to neib_shape in other words, the patches are disjoint. When the step is greater than neib_shape, some elements are omitted. When None, this is the same as neib_shape (patch are disjoint). .. note:: Currently the step size should be chosen in the way that the

    corresponding dimension i (width or height) is equal to n * step\_size_i + neib\_shape_i for some n
  • mode ({‘valid’, ‘ignore_borders’, ‘wrap_centered}) – valid Requires an input that is a multiple of the pooling factor (in each direction). ignore_borders Same as valid, but will ignore the borders if the shape(s) of the input is not a multiple of the pooling factor(s). wrap_centered ?? TODO comment
Returns:

Reshapes the input as a 2D tensor where each row is an pooling example. Pseudo-code of the output:

idx = 0
for i in xrange(list 1 dim):
    for j in xrange(list 2 dim):
        for k in <image column coordinates>:
            for l in <image row coordinates>:
                output[idx,:]
                     = flattened version of ten4[i,j,l:l+r,k:k+c]
                idx += 1

Note

The operation isn’t necessarily implemented internally with these for loops, they’re just the easiest way to describe the output pattern.

Return type:

object

Examples

# Defining variables
images = T.tensor4('images')
neibs = images2neibs(images, neib_shape=(5, 5))

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

# Input tensor (one image 10x10)
im_val = np.arange(100.).reshape((1, 1, 10, 10))

# Function application
neibs_val = window_function(im_val)

Note

The underlying code will construct a 2D tensor of disjoint patches 5x5. The output has shape 4x25.

theano.sandbox.neighbours.neibs2images(neibs, neib_shape, original_shape, mode='valid')

Function neibs2images performs the inverse operation of images2neibs. It inputs the output of images2neibs and reconstructs its input.

Parameters:
  • neibs (matrix) – Like the one obtained by images2neibs.
  • neib_shapeneib_shape that was used in images2neibs.
  • original_shape – Original shape of the 4d tensor given to images2neibs
Returns:

Reconstructs the input of images2neibs, a 4d tensor of shape original_shape.

Return type:

object

Notes

Currently, the function doesn’t support tensors created with neib_step different from default value. This means that it may be impossible to compute the gradient of a variable gained by images2neibs w.r.t. its inputs in this case, because it uses images2neibs for gradient computation.

Examples

Example, which uses a tensor gained in example for images2neibs:

im_new = neibs2images(neibs, (5, 5), im_val.shape)
# Theano function definition
inv_window = theano.function([neibs], im_new)
# Function application
im_new_val = inv_window(neibs_val)

Note

The code will output the initial image array.

class theano.sandbox.neighbours.Images2Neibs(mode='valid')
Parameters:mode ({‘valid’, ‘ignore_borders’, ‘wrap_centered’}) –
‘valid’: Requires an input that is a multiple of the
pooling factor (in each direction).
‘ignore_borders’: Same as valid, but will ignore the borders
if the shape(s) of the input is not a multiple of the pooling factor(s).

‘wrap_centered’ : ?? TODO comment

Returns:Reshapes the input as a 2D tensor where each row is an pooling example.
Return type:object
make_node(ten4, neib_shape, neib_step=None)
Parameters:
  • ten4 (a list of lists of images) – ten4 is of shape (list 1 dim, list 2 dim, row, col).
  • neib_shape – (r,c) where r is the height of the neighborhood in rows and c is the width of the neighborhood in columns.
  • neib_step – (dr,dc) where dr is the number of rows to skip between patch and dc is the number of columns. When None, this is the same as neib_shape (patch are disjoint).
Returns:

A 2D matrix, written using the following pattern idx = 0 for i in xrange(list 1 dim)

for j in xrange(list 2 dim)
for k in <image column coordinates>
for l in <image row coordinates>
output[idx,:]

= flattened version of ten4[i,j,l:l+r,k:k+c]

idx += 1

Note

The op isn’t necessarily implemented internally with these

for loops, they’re just the easiest way to describe the output pattern.

Return type:

matrix