Added an ignore_border mode for the Image2Neibs Op

theano/sandbox/neighbours.py

@@ -12,8 +12,8 @@ if cuda_available:
 
 class Images2Neibs(Op):
     def __init__(self, mode='valid'):
-        if mode not in ['valid','wrap_centered']:
-            raise NotImplementedError("Only the mode valid and wrap_centered have been implemented for the op Images2Neibs")
+        if mode not in ['valid','wrap_centered','ignore_borders']:
+            raise NotImplementedError("Only the mode valid, ignore_borders and wrap_centered have been implemented for the op Images2Neibs")
         self.mode = mode
     def __eq__(self, other):
         return type(self) == type(other) and self.mode==other.mode
@@ -49,13 +49,13 @@ class Images2Neibs(Op):
     def grad(self, inp, grads):
         x, neib_shape, neib_step = inp
         gz, = grads
-        if self.mode=='valid':
-            return [neibs2images(gz, neib_shape, x.shape), None, None]
+        if self.mode in ['valid','ignore_borders']:
+            return [neibs2images(gz, neib_shape, x.shape, mode=self.mode), None, None]
         else:
             raise NotImplementedError()
 
     def c_code_cache_version(self):
-        return (3,)
+        return (4,)
 
     def c_code(self, node, name, inp, out, sub):
         ten4, neib_shape, neib_step = inp
@@ -130,6 +130,9 @@ class Images2Neibs(Op):
             }
             grid_c = 1+(((%(ten4)s->dimensions)[2]-c)/step_x); //number of patch in height
             grid_d = 1+(((%(ten4)s->dimensions)[3]-d)/step_y); //number of patch in width
+        }else if ( "%(mode)s" == "ignore_borders") {
+            grid_c = 1+(((%(ten4)s->dimensions)[2]-c)/step_x); //number of patch in height
+            grid_d = 1+(((%(ten4)s->dimensions)[3]-d)/step_y); //number of patch in width
         }else{
             PyErr_Format(PyExc_TypeError, "Images2Neibs: unknow mode '%(mode)s'");
             %(fail)s;
@@ -221,7 +224,7 @@ class Images2Neibs(Op):
 def images2neibs(ten4, neib_shape, neib_step=None, mode='valid'):
     return Images2Neibs(mode)(ten4, neib_shape, neib_step)
 
-def neibs2images(neibs, neib_shape, original_shape):
+def neibs2images(neibs, neib_shape, original_shape, mode='valid'):
     """
     Inverse of images2neib.
 
@@ -236,8 +239,22 @@ def neibs2images(neibs, neib_shape, original_shape):
     original_shape = T.as_tensor_variable(original_shape)
 
     new_neib_shape = T.stack( original_shape[-1]/neib_shape[1], neib_shape[1] )
-    return images2neibs(neibs.dimshuffle('x','x',0,1), new_neib_shape).reshape(original_shape)
-    #return images2neibs(neibs.reshape((1,1,neibs.shape[0],neibs.shape[1])), new_neib_shape).reshape(original_shape)
+    output_2d = images2neibs(neibs.dimshuffle('x','x',0,1), new_neib_shape, mode=mode)
+    
+    if mode == 'ignore_borders':
+        valid_shape = list(original_shape)
+        valid_shape[2]  = valid_shape[2] / neib_shape[0] * neib_shape[0]
+        valid_shape[3]  = valid_shape[3] / neib_shape[1] * neib_shape[1]
+        output_4d = output_2d.reshape(valid_shape)
+        #padding the borders with zeros
+        for d in [2,3]:
+            pad_shape = list(output_4d.shape)
+            pad_shape[d] = original_shape[d] - valid_shape[d]
+            output_4d = T.concatenate([output_4d,T.zeros(pad_shape)],axis=d)
+    else:
+        output_4d = output_2d.reshape(original_shape)
+    
+    return output_4d
 
 
 # This is work in progress