InterpolateBackprop#

Versioned name: InterpolateBackprop-1

Category: image processing

Short description: Computes the gradients of Interpolate operation.

Attributes:

  • mode

    • Description: specifies type of interpolation

    • Range of values: one of nearest, linear, bilinear, trilinear

    • Type: string

    • Required: yes

  • coordinate_transformation_mode

    • Description: specifies how to transform the coordinate in the resized tensor to the coordinate in the original tensor

    • Range of values: name of the transformation mode in string format (here scale[x] is output_shape[x] / input_shape[x] and x_resized is a coordinate in axis x, for any axis x from the input axes):

      • half_pixel - the coordinate in the original tensor axis x is calculated as ((x_resized + 0.5) / scale[x]) - 0.5.

      • align_corners - the coordinate in the original tensor axis x is calculated as 0 if output_shape[x] == 1 else  x_resized * (input_shape[x] - 1) / (output_shape[x] - 1).

    • Type: string

    • Default value: half_pixel

    • Required: no

  • sizes

    • Description: specifies output shape for spatial axes. sizes and scales can’t be valid at the same time. When sizes is used, optional scales should not be set.

    • Range of values: positive s64 values

    • Type: s64[]

    • Default value: none

    • Required: no

  • scales

    • Description: specifies scales for spatial axes. sizes and scales can’t be valid at the same time. When scales is used, optional size should not be set.

    • Range of values: f32 values

    • Type: f32[]

    • Default value: none

    • Required: no

  • data_format

    • Description: data_format denotes the data format of the input and output data.

    • Range of values: NXC or NCX (X means HW for 2D, DHW for 3D)

    • Type: string

    • Default value: NXC

    • Required: no

Inputs

  • 1: input_forward - original input tensor of Interpolate op. Required.

    • Type: T1

  • 2: output_delta - the gradient tensor with respect to the output. Required.

    • Type: T1

  • 3: sizes - a 1D tensor describing output shape for spatial axes. Optional.

    • Type: T2

Outputs

  • 1: input_delta - the gradient tensor with respect to the input of Interpolate.

    • Type: T1

Types:

  • T1: f32, f16, bf16.

  • T2: s32.

  • Note: The input tensor and the result tensor have the same data type denoted by T1. For example, if input is f32 tensor, then result tensor has f32 data type.