InstanceNormalization

Description

Performs instance normalization.

mean and variance indicate the mean and variance of each channel in each instance.

Input

• x: input tensor of type float16 or float.
• scale: 1D input tensor. The dimension is the same as the length of the C axis of x. The data type is the same as that of input x.
• B: 1D input tensor. The dimension is the same as the length of the C axis of x. The data type is the same as that of input x.

Attribute

epsilon: float. The default value is 1e-05. This avoids division by zero.

Output

y: output tensor of the identical data type and shape to the input.

Constraints

None.

ONNX Opset Support

Opset v8/v9/v10/v11/v12/v13/v14/v15/v16/v17/v18