Method DownsampleFactorMaxGradGrad.grad and method…

theano/tensor/signal/pool.py

@@ -12,7 +12,6 @@ import numpy
 from six import integer_types
 from six.moves import xrange
 import six.moves.builtins as builtins
-
 import theano
 from theano import gof, OpenMPOp, tensor, Variable, Apply
 
@@ -995,6 +994,15 @@ class DownsampleFactorMaxGradGrad(OpenMPOp):
     def infer_shape(self, node, in_shapes):
         return [in_shapes[1]]
 
+    def grad(self, inp, grads):
+        x, maxout, ggx = inp
+        gz, = grads
+        return [theano.tensor.zeros_like(x),
+                theano.tensor.zeros_like(maxout),
+                MaxPoolGrad(
+                    self.ds, ignore_border=self.ignore_border,
+                    st=self.st, padding=self.padding)(x, maxout, gz)]
+
     def c_code(self, node, name, inp, out, sub):
         if self.mode != 'max':
             raise theano.gof.utils.MethodNotDefined()

theano/tensor/signal/tests/test_pool.py

@@ -11,7 +11,8 @@ import theano.tensor as tensor
 from theano.tests import unittest_tools as utt
 from theano.tensor.signal.pool import (Pool, pool_2d,
                                        MaxPoolGrad, AveragePoolGrad,
-                                       max_pool_2d_same_size)
+                                       max_pool_2d_same_size,
+                                       DownsampleFactorMaxGradGrad)
 
 from theano.tensor.signal.downsample import DownsampleFactorMaxGrad
 
@@ -652,6 +653,35 @@ class TestDownsampleFactorMax(utt.InferShapeTester):
 
         assert numpy.allclose(fn_hess([1, 2]), [[0., 0.], [0., 982.7667]])
 
+    def test_DownsampleFactorMaxGradGrad_grad(self):
+        rng = numpy.random.RandomState(utt.fetch_seed())
+        imgsizes = ((10, 10), (10, 5), (5, 5))
+        maxpoolsizes = ((5, 3), (3, 5), (3, 3))
+        stridesizes = ((3, 2), (2, 3), (3, 3))
+        paddingsizes = ((2, 2), (2, 1), (2, 2))
+
+        for i in range(len(imgsizes)):
+            imgsize = imgsizes[i]
+            imval1 = rng.rand(1, 1, imgsize[0], imgsize[1]) * 10.0
+            imval2 = rng.rand(1, 1, imgsize[0], imgsize[1]) * 10.0
+            maxpoolsize = maxpoolsizes[i]
+            stridesize = stridesizes[i]
+            paddingsize = paddingsizes[i]
+
+            def mp(input1, input2):
+                pooled_out = Pool(
+                    maxpoolsize, ignore_border=True,
+                    st=stridesize,
+                    padding=paddingsize,
+                    )(input1)
+                out = DownsampleFactorMaxGradGrad(
+                    ds=maxpoolsize,
+                    ignore_border=True,
+                    st=stridesize,
+                    padding=paddingsize)(input1, pooled_out, input2)
+                return out
+            utt.verify_grad(mp, [imval1, imval2], rng=rng)
+
     def test_max_pool_2d_2D(self):
         rng = numpy.random.RandomState(utt.fetch_seed())
         maxpoolshps = ((1, 1), (3, 2))