Edit

DML_CONVOLUTION_INTEGER_OPERATOR_DESC structure (directml.h)

  • Article

Performs a convolution of the FilterTensor with the InputTensor. This operator performs forward convolution on integer data. Optional zero point tensors can also be used to subtract zero point values from the input and filter tensor.

Syntax

struct DML_CONVOLUTION_INTEGER_OPERATOR_DESC {
  const DML_TENSOR_DESC *InputTensor;
  const DML_TENSOR_DESC *InputZeroPointTensor;
  const DML_TENSOR_DESC *FilterTensor;
  const DML_TENSOR_DESC *FilterZeroPointTensor;
  const DML_TENSOR_DESC *OutputTensor;
  UINT                  DimensionCount;
  const UINT            *Strides;
  const UINT            *Dilations;
  const UINT            *StartPadding;
  const UINT            *EndPadding;
  UINT                  GroupCount;
};

Members

InputTensor

Type: const DML_TENSOR_DESC*

A tensor containing the input data. The expected dimensions of the InputTensor are { BatchCount, InputChannelCount, InputHeight, InputWidth }.

InputZeroPointTensor

Type: _Maybenull_ const DML_TENSOR_DESC*

An optional tensor containing the input zero point data. The expected dimensions of the InputZeroPointTensor are { 1, 1, 1, 1 }.

FilterTensor

Type: const DML_TENSOR_DESC*

A tensor containing the filter data. The expected dimensions of the FilterTensor are { FilterBatchCount, FilterChannelCount, FilterHeight, FilterWidth }.

FilterZeroPointTensor

Type: _Maybenull_ const DML_TENSOR_DESC*

An optional tensor containing the filter zero point data. The expected dimensions of the FilterZeroPointTensor are { 1, 1, 1, 1 } if per tensor quantization is required, or { 1, OutputChannelCount, 1, 1 } if per-channel quantization is required.

OutputTensor

Type: const DML_TENSOR_DESC*

The tensor to write the results to. The expected dimensions of the OutputTensor are { BatchCount, OutputChannelCount, OutputHeight, OutputWidth }.

DimensionCount

Type: UINT

The number of spatial dimensions for the convolution operation. Spatial dimensions are the lower dimensions of the convolution FilterTensor. This value also determines the size of the Strides, Dilations, StartPadding, and EndPadding arrays. Only a value of 2 is supported.

Strides

Type: _Field_size_(DimensionCount) const UINT*

An array containing the strides of the convolution operation. These strides are applied to the convolution filter. They are separate from the tensor strides included in DML_TENSOR_DESC.

Dilations

Type: _Field_size_(DimensionCount) const UINT*

An array containing the dilations of the convolution operation. Dilations are strides applied to the elements of the filter kernel. This has the effect of simulating a larger filter kernel by padding the internal filter kernel elements with zeros.

StartPadding

Type: _Field_size_(DimensionCount) const UINT*

An array containing the padding values to be applied to the beginning of each spatial dimension of the filter and input tensor of the convolution operation.

EndPadding

Type: _Field_size_(DimensionCount) const UINT*

An array containing the padding values to be applied to the end of each spatial dimension of the filter and input tensor of the convolution operation.

GroupCount

Type: UINT

The number of groups which to divide the convolution operation up into. GroupCount can be used to achieve depth-wise convolution by setting the GroupCount equal to the input channel count. This divides the convolution up into a separate convolution per input channel.

Availability

This operator was introduced in DML_FEATURE_LEVEL_2_1.

Tensor constraints

  • FilterZeroPointTensor and InputZeroPointTensor must have the same DimensionCount.
  • FilterTensor, InputTensor, and OutputTensor must have the same DimensionCount.
  • InputTensor and InputZeroPointTensor must have the same DataType.
  • FilterTensor and FilterZeroPointTensor must have the same DataType.

Tensor support

DML_FEATURE_LEVEL_4_0 and above

Tensor Kind Dimensions Supported dimension counts Supported data types
InputTensor Input { BatchCount, InputChannelCount, [InputHeight], InputWidth } 3 to 4 INT8, UINT8
InputZeroPointTensor Optional input { [1], [1], [1], 1 } 1 to 4 INT8, UINT8
FilterTensor Input { FilterBatchCount, FilterChannelCount, [FilterHeight], FilterWidth } 3 to 4 INT8, UINT8
FilterZeroPointTensor Optional input { [1], FilterZeroPointChannelCount, [1], [1] } 1 to 4 INT8, UINT8
OutputTensor Output { BatchCount, OutputChannelCount, [OutputHeight], OutputWidth } 3 to 4 INT32

DML_FEATURE_LEVEL_2_1 and above

Tensor Kind Dimensions Supported dimension counts Supported data types
InputTensor Input { BatchCount, InputChannelCount, InputHeight, InputWidth } 4 INT8, UINT8
InputZeroPointTensor Optional input { 1, 1, 1, 1 } 4 INT8, UINT8
FilterTensor Input { FilterBatchCount, FilterChannelCount, FilterHeight, FilterWidth } 4 INT8, UINT8
FilterZeroPointTensor Optional input { 1, FilterZeroPointChannelCount, 1, 1 } 4 INT8, UINT8
OutputTensor Output { BatchCount, OutputChannelCount, OutputHeight, OutputWidth } 4 INT32

Requirements

Requirement Value
Minimum supported client Windows 10 Build 20348
Minimum supported server Windows 10 Build 20348
Header directml.h