chainer.functions.unpooling_nd¶
-
chainer.functions.
unpooling_nd
(x, ksize, stride=None, pad=0, outsize=None, cover_all=True)[source]¶ Inverse operation of N-dimensional spatial pooling.
Warning
This feature is experimental. The interface can change in the future.
This function acts similarly to
DeconvolutionND
, but it spreads input N-dimensional array’s value without any parameter instead of computing the inner products.Parameters: - x (Variable) – Input variable.
- ksize (int or pair of ints) – Size of pooling window
\((k_1, k_2, ..., k_N)\).
ksize=k
is equivalent to(k, k, ..., k)
. - stride (int, pair of ints or None) – Stride of pooling applications
\((s_1, s_2, ..., s_N)\).
stride=s
is equivalent to(s, s, ..., s)
. IfNone
is specified, then it uses same stride as the pooling window size. - pad (int or pair of ints) – Spatial padding width for the input array
\((p_1, p_2, ..., p_N)\).
pad=p
is equivalent to(p, p, ..., p)
. - outsize (None or pair of ints) – Expected output size of unpooling
operation \((out_1, out_2, ..., out_N)\). If
None
, the size is estimated from input size, stride and padding. - cover_all (bool) – If
True
, the pooling window is assumed to cover all of the output array, eventually the output size may be smaller than that in the casecover_all
isFalse
.
Returns: Output variable.
Return type: