MeanVarianceNormalization#

MeanVarianceNormalization - 13#

Version

This version of the operator has been available since version 13.

Summary

A MeanVarianceNormalization Function: Perform mean variance normalization on the input tensor X using formula: <br/> ` (X-EX)/sqrt(E(X-EX)^2) `

Attributes

  • axes: A list of integers, along which to reduce. The default is to caculate along axes [0,2,3] for calculating mean and variance along each channel. Two variables with the same C-coordinate are associated with the same mean and variance. Default value is [0 2 3].

Inputs

  • X (heterogeneous) - T: Input tensor

Outputs

  • Y (heterogeneous) - T: Output tensor

Type Constraints

  • T in ( tensor(bfloat16), tensor(double), tensor(float), tensor(float16) ): Constrain input and output types to all numeric tensors.

Examples

default

node = onnx.helper.make_node(
    'MeanVarianceNormalization',
    inputs=['X'],
    outputs=['Y']
)

input_data = np.array([[[[0.8439683], [0.5665144], [0.05836735]],
    [[0.02916367], [0.12964272], [0.5060197]],
    [[0.79538304], [0.9411346], [0.9546573]]],
    [[[0.17730942], [0.46192095], [0.26480448]],
    [[0.6746842], [0.01665257], [0.62473077]],
    [[0.9240844], [0.9722341], [0.11965699]]],
    [[[0.41356155], [0.9129373], [0.59330076]],
    [[0.81929934], [0.7862604], [0.11799799]],
    [[0.69248444], [0.54119414], [0.07513223]]]], dtype=np.float32)

# Calculate expected output data
data_mean = np.mean(input_data, axis=(0, 2, 3), keepdims=1)
data_mean_squared = np.power(data_mean, 2)
data_squared = np.power(input_data, 2)
data_squared_mean = np.mean(data_squared, axis=(0, 2, 3), keepdims=1)
std = np.sqrt(data_squared_mean - data_mean_squared)
expected_output = (input_data - data_mean) / (std + 1e-9)

expect(node, inputs=[input_data], outputs=[expected_output],
       name='test_mvn')

Differences

00A MeanVarianceNormalization Function: Perform mean variance normalizationA MeanVarianceNormalization Function: Perform mean variance normalization
11on the input tensor X using formula:
(X-EX)/sqrt(E(X-EX)^2)
on the input tensor X using formula:
(X-EX)/sqrt(E(X-EX)^2)
22
33**Attributes****Attributes**
44
55* **axes**:* **axes**:
66 A list of integers, along which to reduce. The default is to A list of integers, along which to reduce. The default is to
77 caculate along axes [0,2,3] for calculating mean and variance along caculate along axes [0,2,3] for calculating mean and variance along
88 each channel. Two variables with the same C-coordinate are each channel. Two variables with the same C-coordinate are
99 associated with the same mean and variance. Default value is [0 2 3]. associated with the same mean and variance. Default value is [0 2 3].
1010
1111**Inputs****Inputs**
1212
1313* **X** (heterogeneous) - **T**:* **X** (heterogeneous) - **T**:
1414 Input tensor Input tensor
1515
1616**Outputs****Outputs**
1717
1818* **Y** (heterogeneous) - **T**:* **Y** (heterogeneous) - **T**:
1919 Output tensor Output tensor
2020
2121**Type Constraints****Type Constraints**
2222
2323* **T** in (* **T** in (
24 tensor(bfloat16),
2425 tensor(double), tensor(double),
2526 tensor(float), tensor(float),
2627 tensor(float16) tensor(float16)
2728 ): ):
2829 Constrain input and output types to all numeric tensors. Constrain input and output types to all numeric tensors.

MeanVarianceNormalization - 9#

Version

This version of the operator has been available since version 9.

Summary

A MeanVarianceNormalization Function: Perform mean variance normalization on the input tensor X using formula: <br/> ` (X-EX)/sqrt(E(X-EX)^2) `

Attributes

  • axes: A list of integers, along which to reduce. The default is to caculate along axes [0,2,3] for calculating mean and variance along each channel. Two variables with the same C-coordinate are associated with the same mean and variance. Default value is [0 2 3].

Inputs

  • X (heterogeneous) - T: Input tensor

Outputs

  • Y (heterogeneous) - T: Output tensor

Type Constraints

  • T in ( tensor(double), tensor(float), tensor(float16) ): Constrain input and output types to all numeric tensors.