Add default values, fix imports, fix documentation

python/mxnet/gluon/data/vision/transforms.py

@@ -178,7 +178,7 @@ class Normalize(HybridBlock):
       [-1.4711051  -1.5200229 ]]]
     <NDArray 3x4x2 @cpu(0)>
     """
-    def __init__(self, mean, std):
+    def __init__(self, mean=0.0, std=1.0):
         super(Normalize, self).__init__()
         self._mean = mean
         self._std = std

src/operator/image/normalize_op-inl.h

@@ -26,15 +26,7 @@
 #define MXNET_OPERATOR_IMAGE_NORMALIZE_OP_INL_H_
 
 
-#include <mxnet/base.h>
-#include <algorithm>
 #include <vector>
-#include <cmath>
-#include <limits>
-#include <algorithm>
-#include <utility>
-#include "../mxnet_op.h"
-#include "../operator_common.h"
 #include "../elemwise_op_common.h"
 
 namespace mxnet {
@@ -44,12 +36,19 @@ namespace image {
 // Parameter registration for image Normalize operator
 struct NormalizeParam : public dmlc::Parameter<NormalizeParam> {
   nnvm::Tuple<float> mean;
+  nnvm::Tuple<float> default_mean = {0.0f, 0.0f, 0.0f, 0.0f};
   nnvm::Tuple<float> std;
+  nnvm::Tuple<float> default_std = {1.0f, 1.0f, 1.0f, 1.0f};
+
   DMLC_DECLARE_PARAMETER(NormalizeParam) {
     DMLC_DECLARE_FIELD(mean)
-    .describe("Sequence of means for each channel.");
+    .set_default(default_mean)
+    .describe("Sequence of means for each channel. "
+              "Default value is 0.");
     DMLC_DECLARE_FIELD(std)
-    .describe("Sequence of standard deviations for each channel.");
+    .set_default(default_std)
+    .describe("Sequence of standard deviations for each channel. "
+              "Default value is 1.");
   }
 };
 

src/operator/image/normalize_op.cc

@@ -33,14 +33,17 @@ NNVM_REGISTER_OP(_image_normalize)
 .describe(R"code(Normalize an tensor of shape (C x H x W) or (N x C x H x W) with mean and
     standard deviation.
 
-    Given mean `(m1, ..., mn)` and std `(s1, ..., sn)` for `n` channels,
+    Given mean `(m1, ..., mn)` and std `(s\ :sub:`1`\ , ..., s\ :sub:`n`)` for `n` channels,
     this transform normalizes each channel of the input tensor with:
+
 .. math::
-        
-        output[i] = (input[i] - mi) / si
-    
+
+        output[i] = (input[i] - m\ :sub:`i`\ ) / s\ :sub:`i`
+
     If mean or std is scalar, the same value will be applied to all channels.
 
+    Default value for mean is 0.0 and stand deviation is 1.0.
+
 Example:
 
     .. code-block:: python

tests/python/gpu/test_gluon_transforms.py

@@ -55,6 +55,12 @@ def test_normalize():
     data_expected_4d[1][:][:][2] = data_expected_4d[1][:][:][2] - 2.0
     assert_almost_equal(data_expected_4d, out_nd_4d.asnumpy())
 
+    # Default normalize values i.e., mean=0, std=1
+    data_in_3d_def = nd.random.uniform(0, 1, (3, 300, 300))
+    out_nd_3d_def = transforms.Normalize()(data_in_3d_def)
+    data_expected_3d_def = data_in_3d.asnumpy()
+    assert_almost_equal(data_expected_3d_def, out_nd_3d_def.asnumpy())
+
     # Invalid Input - Neither 3D or 4D input
     invalid_data_in = nd.random.uniform(0, 1, (5, 5, 3, 300, 300))
     normalize_transformer = transforms.Normalize(mean=(0, 1, 2), std=(3, 2, 1))