oneAPI Deep Neural Network Library (oneDNN)
Performance library for Deep Learning
1.96.0
Deconvolution

A primitive to perform 1D, 2D or 3D deconvolution. More...

Classes

struct  dnnl::deconvolution_forward
 Deconvolution forward propagation primitive. More...
 
struct  dnnl::deconvolution_backward_data
 Deconvolution backward propagation primitive. More...
 
struct  dnnl::deconvolution_backward_weights
 Deconvolution weights gradient primitive. More...
 

Typedefs

typedef dnnl_convolution_desc_t dnnl_deconvolution_desc_t
 A descriptor of a deconvolution operation.
 

Functions

dnnl_status_t DNNL_API dnnl_deconvolution_forward_desc_init (dnnl_deconvolution_desc_t *deconv_desc, dnnl_prop_kind_t prop_kind, dnnl_alg_kind_t alg_kind, const dnnl_memory_desc_t *src_desc, const dnnl_memory_desc_t *weights_desc, const dnnl_memory_desc_t *bias_desc, const dnnl_memory_desc_t *dst_desc, const dnnl_dims_t strides, const dnnl_dims_t padding_l, const dnnl_dims_t padding_r)
 Initializes a descriptor for a deconvolution forward propagation primitive. More...
 
dnnl_status_t DNNL_API dnnl_dilated_deconvolution_forward_desc_init (dnnl_deconvolution_desc_t *deconv_desc, dnnl_prop_kind_t prop_kind, dnnl_alg_kind_t alg_kind, const dnnl_memory_desc_t *src_desc, const dnnl_memory_desc_t *weights_desc, const dnnl_memory_desc_t *bias_desc, const dnnl_memory_desc_t *dst_desc, const dnnl_dims_t strides, const dnnl_dims_t dilates, const dnnl_dims_t padding_l, const dnnl_dims_t padding_r)
 Initializes a descriptor for a dilated deconvolution forward propagation primitive. More...
 
dnnl_status_t DNNL_API dnnl_deconvolution_backward_data_desc_init (dnnl_deconvolution_desc_t *deconv_desc, dnnl_alg_kind_t alg_kind, const dnnl_memory_desc_t *diff_src_desc, const dnnl_memory_desc_t *weights_desc, const dnnl_memory_desc_t *diff_dst_desc, const dnnl_dims_t strides, const dnnl_dims_t padding_l, const dnnl_dims_t padding_r)
 Initializes a descriptor for a deconvolution backward propagation primitive. More...
 
dnnl_status_t DNNL_API dnnl_dilated_deconvolution_backward_data_desc_init (dnnl_deconvolution_desc_t *deconv_desc, dnnl_alg_kind_t alg_kind, const dnnl_memory_desc_t *diff_src_desc, const dnnl_memory_desc_t *weights_desc, const dnnl_memory_desc_t *diff_dst_desc, const dnnl_dims_t strides, const dnnl_dims_t dilates, const dnnl_dims_t padding_l, const dnnl_dims_t padding_r)
 Initializes a descriptor for a dilated deconvolution backward propagation primitive. More...
 
dnnl_status_t DNNL_API dnnl_deconvolution_backward_weights_desc_init (dnnl_deconvolution_desc_t *deconv_desc, dnnl_alg_kind_t alg_kind, const dnnl_memory_desc_t *src_desc, const dnnl_memory_desc_t *diff_weights_desc, const dnnl_memory_desc_t *diff_bias_desc, const dnnl_memory_desc_t *diff_dst_desc, const dnnl_dims_t strides, const dnnl_dims_t padding_l, const dnnl_dims_t padding_r)
 Initializes a descriptor for a deconvolution weights gradient primitive. More...
 
dnnl_status_t DNNL_API dnnl_dilated_deconvolution_backward_weights_desc_init (dnnl_deconvolution_desc_t *deconv_desc, dnnl_alg_kind_t alg_kind, const dnnl_memory_desc_t *src_desc, const dnnl_memory_desc_t *diff_weights_desc, const dnnl_memory_desc_t *diff_bias_desc, const dnnl_memory_desc_t *diff_dst_desc, const dnnl_dims_t strides, const dnnl_dims_t dilates, const dnnl_dims_t padding_l, const dnnl_dims_t padding_r)
 Initializes a descriptor for a dilated deconvolution weights gradient primitive. More...
 

Detailed Description

A primitive to perform 1D, 2D or 3D deconvolution.

Supported variants are forward propagation, backward propagation, and weights gradient with or without bias.

Function Documentation

◆ dnnl_deconvolution_forward_desc_init()

dnnl_status_t DNNL_API dnnl_deconvolution_forward_desc_init ( dnnl_deconvolution_desc_t deconv_desc,
dnnl_prop_kind_t  prop_kind,
dnnl_alg_kind_t  alg_kind,
const dnnl_memory_desc_t src_desc,
const dnnl_memory_desc_t weights_desc,
const dnnl_memory_desc_t bias_desc,
const dnnl_memory_desc_t dst_desc,
const dnnl_dims_t  strides,
const dnnl_dims_t  padding_l,
const dnnl_dims_t  padding_r 
)

Initializes a descriptor for a deconvolution forward propagation primitive.

Note
Memory descriptors can be initialized with dnnl_format_tag_any or with format_kind set to dnnl_format_kind_any.

Arrays strides, padding_l, and padding_r contain values for spatial dimensions only and hence must have the same number of elements as there are spatial dimensions. The order of values is the same as in the tensor: depth (for 3D tensors), height (for 3D and 2D tensors), and width.

Parameters
deconv_descOutput descriptor for a deconvolution primitive.
prop_kindPropagation kind. Possible values are dnnl_forward_training and dnnl_forward_inference.
alg_kindDeconvolution algorithm. Possible values are dnnl_deconvolution_direct, dnnl_deconvolution_winograd.
src_descSource memory descriptor.
weights_descWeights memory descriptor.
bias_descBias memory descriptor. Passing NULL, a zero memory descriptor, or a memory descriptor with format_kind set to dnnl_format_kind_undef disables the bias term.
dst_descDestination memory descriptor.
stridesArray of strides for spatial dimension.
padding_lArray of padding values for low indices for each spatial dimension ([[front,] top,] left).
padding_rArray of padding values for high indices for each spatial dimension ([[back,] bottom,] right). Can be NULL in which case padding is considered to be symmetrical.
Returns
dnnl_success on success and a status describing the error otherwise.

◆ dnnl_dilated_deconvolution_forward_desc_init()

dnnl_status_t DNNL_API dnnl_dilated_deconvolution_forward_desc_init ( dnnl_deconvolution_desc_t deconv_desc,
dnnl_prop_kind_t  prop_kind,
dnnl_alg_kind_t  alg_kind,
const dnnl_memory_desc_t src_desc,
const dnnl_memory_desc_t weights_desc,
const dnnl_memory_desc_t bias_desc,
const dnnl_memory_desc_t dst_desc,
const dnnl_dims_t  strides,
const dnnl_dims_t  dilates,
const dnnl_dims_t  padding_l,
const dnnl_dims_t  padding_r 
)

Initializes a descriptor for a dilated deconvolution forward propagation primitive.

Note
Memory descriptors can be initialized with dnnl_format_tag_any or with format_kind set to dnnl_format_kind_any.

Arrays strides, dilates, padding_l, and padding_r contain values for spatial dimensions only and hence must have the same number of elements as there are spatial dimensions. The order of values is the same as in the tensor: depth (for 3D tensors), height (for 3D and 2D tensors), and width.

Parameters
deconv_descOutput descriptor for a deconvolution primitive.
prop_kindPropagation kind. Possible values are dnnl_forward_training and dnnl_forward_inference.
alg_kindDeconvolution algorithm. Possible values are dnnl_deconvolution_direct, dnnl_deconvolution_winograd.
src_descSource memory descriptor.
weights_descWeights memory descriptor.
bias_descBias memory descriptor. Passing NULL, a zero memory descriptor, or a memory descriptor with format_kind set to dnnl_format_kind_undef disables the bias term.
dst_descDestination memory descriptor.
stridesArray of strides for spatial dimension.
dilatesArray of dilations for spatial dimension. A zero value means no dilation in the corresponding dimension.
padding_lArray of padding values for low indices for each spatial dimension ([[front,] top,] left).
padding_rArray of padding values for high indices for each spatial dimension ([[back,] bottom,] right). Can be NULL in which case padding is considered to be symmetrical.
Returns
dnnl_success on success and a status describing the error otherwise.

◆ dnnl_deconvolution_backward_data_desc_init()

dnnl_status_t DNNL_API dnnl_deconvolution_backward_data_desc_init ( dnnl_deconvolution_desc_t deconv_desc,
dnnl_alg_kind_t  alg_kind,
const dnnl_memory_desc_t diff_src_desc,
const dnnl_memory_desc_t weights_desc,
const dnnl_memory_desc_t diff_dst_desc,
const dnnl_dims_t  strides,
const dnnl_dims_t  padding_l,
const dnnl_dims_t  padding_r 
)

Initializes a descriptor for a deconvolution backward propagation primitive.

Note
Memory descriptors can be initialized with dnnl_format_tag_any or with format_kind set to dnnl_format_kind_any.

Arrays strides, padding_l, and padding_r contain values for spatial dimensions only and hence must have the same number of elements as there are spatial dimensions. The order of values is the same as in the tensor: depth (for 3D tensors), height (for 3D and 2D tensors), and width.

Parameters
deconv_descOutput descriptor for a deconvolution primitive.
alg_kindDeconvolution algorithm. Possible values are dnnl_deconvolution_direct, dnnl_deconvolution_winograd.
diff_src_descDiff source memory descriptor.
weights_descWeights memory descriptor.
diff_dst_descDiff destination memory descriptor.
stridesArray of strides for spatial dimension.
padding_lArray of padding values for low indices for each spatial dimension ([[front,] top,] left).
padding_rArray of padding values for high indices for each spatial dimension ([[back,] bottom,] right). Can be NULL in which case padding is considered to be symmetrical.
Returns
dnnl_success on success and a status describing the error otherwise.

◆ dnnl_dilated_deconvolution_backward_data_desc_init()

dnnl_status_t DNNL_API dnnl_dilated_deconvolution_backward_data_desc_init ( dnnl_deconvolution_desc_t deconv_desc,
dnnl_alg_kind_t  alg_kind,
const dnnl_memory_desc_t diff_src_desc,
const dnnl_memory_desc_t weights_desc,
const dnnl_memory_desc_t diff_dst_desc,
const dnnl_dims_t  strides,
const dnnl_dims_t  dilates,
const dnnl_dims_t  padding_l,
const dnnl_dims_t  padding_r 
)

Initializes a descriptor for a dilated deconvolution backward propagation primitive.

Note
Memory descriptors can be initialized with dnnl_format_tag_any or with format_kind set to dnnl_format_kind_any.

Arrays strides, dilates, padding_l, and padding_r contain values for spatial dimensions only and hence must have the same number of elements as there are spatial dimensions. The order of values is the same as in the tensor: depth (for 3D tensors), height (for 3D and 2D tensors), and width.

Parameters
deconv_descOutput descriptor for a deconvolution primitive.
alg_kindDeconvolution algorithm. Possible values are dnnl_deconvolution_direct, dnnl_deconvolution_winograd.
diff_src_descDiff source memory descriptor.
weights_descWeights memory descriptor.
diff_dst_descDiff destination memory descriptor.
stridesArray of strides for spatial dimension.
dilatesArray of dilations for spatial dimension. A zero value means no dilation in the corresponding dimension.
padding_lArray of padding values for low indices for each spatial dimension ([[front,] top,] left).
padding_rArray of padding values for high indices for each spatial dimension ([[back,] bottom,] right). Can be NULL in which case padding is considered to be symmetrical.
Returns
dnnl_success on success and a status describing the error otherwise.

◆ dnnl_deconvolution_backward_weights_desc_init()

dnnl_status_t DNNL_API dnnl_deconvolution_backward_weights_desc_init ( dnnl_deconvolution_desc_t deconv_desc,
dnnl_alg_kind_t  alg_kind,
const dnnl_memory_desc_t src_desc,
const dnnl_memory_desc_t diff_weights_desc,
const dnnl_memory_desc_t diff_bias_desc,
const dnnl_memory_desc_t diff_dst_desc,
const dnnl_dims_t  strides,
const dnnl_dims_t  padding_l,
const dnnl_dims_t  padding_r 
)

Initializes a descriptor for a deconvolution weights gradient primitive.

Note
Memory descriptors can be initialized with dnnl_format_tag_any or with format_kind set to dnnl_format_kind_any.

Arrays strides, padding_l, and padding_r contain values for spatial dimensions only and hence must have the same number of elements as there are spatial dimensions. The order of values is the same as in the tensor: depth (for 3D tensors), height (for 3D and 2D tensors), and width.

Parameters
deconv_descOutput descriptor for a deconvolution primitive.
alg_kindDeconvolution algorithm. Possible values are dnnl_deconvolution_direct, dnnl_deconvolution_winograd.
src_descSource memory descriptor.
diff_weights_descDiff weights memory descriptor.
diff_bias_descDiff bias memory descriptor. Passing NULL, a zero memory descriptor, or a memory descriptor with format_kind set to dnnl_format_kind_undef disables the bias term.
diff_dst_descDiff destination memory descriptor.
stridesArray of strides for spatial dimension.
padding_lArray of padding values for low indices for each spatial dimension ([[front,] top,] left).
padding_rArray of padding values for high indices for each spatial dimension ([[back,] bottom,] right). Can be NULL in which case padding is considered to be symmetrical.
Returns
dnnl_success on success and a status describing the error otherwise.

◆ dnnl_dilated_deconvolution_backward_weights_desc_init()

dnnl_status_t DNNL_API dnnl_dilated_deconvolution_backward_weights_desc_init ( dnnl_deconvolution_desc_t deconv_desc,
dnnl_alg_kind_t  alg_kind,
const dnnl_memory_desc_t src_desc,
const dnnl_memory_desc_t diff_weights_desc,
const dnnl_memory_desc_t diff_bias_desc,
const dnnl_memory_desc_t diff_dst_desc,
const dnnl_dims_t  strides,
const dnnl_dims_t  dilates,
const dnnl_dims_t  padding_l,
const dnnl_dims_t  padding_r 
)

Initializes a descriptor for a dilated deconvolution weights gradient primitive.

Note
Memory descriptors can be initialized with dnnl_format_tag_any or with format_kind set to dnnl_format_kind_any.

Arrays strides, dilates, padding_l, and padding_r contain values for spatial dimensions only and hence must have the same number of elements as there are spatial dimensions. The order of values is the same as in the tensor: depth (for 3D tensors), height (for 3D and 2D tensors), and width.

Parameters
deconv_descOutput descriptor for a deconvolution primitive.
alg_kindDeconvolution algorithm. Possible values are dnnl_deconvolution_direct, dnnl_deconvolution_winograd.
src_descSource memory descriptor.
diff_weights_descDiff weights memory descriptor.
diff_bias_descDiff bias memory descriptor. Passing NULL, a zero memory descriptor, or a memory descriptor with format_kind set to dnnl_format_kind_undef disables the bias term.
diff_dst_descDiff destination memory descriptor.
stridesArray of strides for spatial dimension.
dilatesArray of dilations for spatial dimension. A zero value means no dilation in the corresponding dimension.
padding_lArray of padding values for low indices for each spatial dimension ([[front,] top,] left).
padding_rArray of padding values for high indices for each spatial dimension ([[back,] bottom,] right). Can be NULL in which case padding is considered to be symmetrical.
Returns
dnnl_success on success and a status describing the error otherwise.