clDNN
cldnn::pooling Struct Reference

Performs "pooling" operation which is a form of non-linear down-sampling. More...

#include <pooling.hpp>

Inheritance diagram for cldnn::pooling:
Collaboration diagram for cldnn::pooling:

Public Types

typedef cldnn_pooling_desc dto
 

Public Member Functions

 pooling (const primitive_id &id, const primitive_id &input, pooling_mode mode, const tensor &size, const tensor &stride, const tensor &input_offset={ 0, 0, 0, 0 }, const padding &output_padding=padding())
 Constructs pooling primitive. More...
 
 pooling (const primitive_id &id, const primitive_id &input, pooling_mode mode, const tensor &size, const tensor &stride, const tensor &input_offset, tensor output_size, const padding &output_padding=padding())
 Constructs pooling primitive (computes input paddings to match output size). More...
 
 pooling (const dto *dto)
 Constructs a copy from C API cldnn_pooling_desc.
 
- Public Member Functions inherited from cldnn::primitive_base< pooling, cldnn_pooling_desc >
const cldnn_primitive_descget_dto () const override
 Returns pointer to a C API primitive descriptor casted to cldnn_primitive_desc.
 
- Public Member Functions inherited from cldnn::primitive
 primitive (const primitive_type_id &type, const primitive_id &id, const std::vector< primitive_id > &input, const padding &output_padding=padding())
 
 primitive (const cldnn_primitive_desc *dto)
 Constructs a copy from basic C API cldnn_primitive_desc.
 
std::vector< std::reference_wrapper< primitive_id > > dependecies ()
 Returns references to all primitive ids on which this primitive depends - inputs, weights, biases, etc.
 
std::vector< primitive_iddependecies () const
 Returns copy of all primitive ids on which this primitive depends - inputs, weights, biases, etc.
 
 operator primitive_id () const
 Implicit conversion to primiitive id.
 

Static Public Member Functions

static primitive_type_id type_id ()
 
static pooling create_with_output_size (const primitive_id &id, const primitive_id &input, tensor output_size, pooling_mode mode, const tensor &size, const tensor &stride, const tensor &input_offset={ 0, 0, 0, 0 }, const padding &output_padding=padding())
 Constructs pooling primitive (computes input paddings to match output size). More...
 

Public Attributes

pooling_mode mode
 Pooling mode.
 
tensor input_offset
 Defines a shift, relative to (0,0) position of the input buffer, where (0,0) point of the pooling window should start calculations.
 
tensor stride
 Defines shift in input buffer between adjacent calculations of output values.
 
tensor size
 Pooling kernel size.
 
bool with_output_size
 Indicates that the primitive has user-defined output size (non-zero value).
 
tensor output_size
 User-defined output data size of the primitive (w/o padding).
 
- Public Attributes inherited from cldnn::primitive
const primitive_type_id type
 Primitive's type id.
 
const primitive_id id
 Primitive's id.
 
fixed_size_vector_ref input
 List of ids of input primitives.
 
padding output_padding
 Requested output padding.
 

Protected Member Functions

void update_dto (dto &dto) const override
 
- Protected Member Functions inherited from cldnn::primitive_base< pooling, cldnn_pooling_desc >
 primitive_base (const primitive_id &id, const std::vector< primitive_id > &input, const padding &output_padding=padding())
 
 primitive_base (const cldnn_pooling_desc *dto)
 
- Protected Member Functions inherited from cldnn::primitive
virtual std::vector< std::reference_wrapper< const primitive_id > > get_dependencies () const
 

Additional Inherited Members

- Protected Attributes inherited from cldnn::primitive
primitive_id_arr _input
 

Detailed Description

Performs "pooling" operation which is a form of non-linear down-sampling.

Pools the input image by taking the max, average, etc. within regions.

Definition at line 44 of file pooling.hpp.

Constructor & Destructor Documentation

◆ pooling() [1/2]

cldnn::pooling::pooling ( const primitive_id id,
const primitive_id input,
pooling_mode  mode,
const tensor size,
const tensor stride,
const tensor input_offset = { 0,0,0,0 },
const padding output_padding = padding() 
)
inline

Constructs pooling primitive.

Parameters
idThis primitive id.
inputInput primitive id.
modePooling mode.
strideDefines shift in input buffer between adjacent calculations of output values.
sizePooling kernel size.

Definition at line 54 of file pooling.hpp.

◆ pooling() [2/2]

cldnn::pooling::pooling ( const primitive_id id,
const primitive_id input,
pooling_mode  mode,
const tensor size,
const tensor stride,
const tensor input_offset,
tensor  output_size,
const padding output_padding = padding() 
)
inline

Constructs pooling primitive (computes input paddings to match output size).

Parameters
idThis primitive id.
inputInput primitive id.
modePooling mode.
strideDefines shift in input buffer between adjacent calculations of output values.
sizePooling kernel size.
output_sizeUser-defined output data size of the primitive (w/o padding).

Definition at line 78 of file pooling.hpp.

Member Function Documentation

◆ create_with_output_size()

static pooling cldnn::pooling::create_with_output_size ( const primitive_id id,
const primitive_id input,
tensor  output_size,
pooling_mode  mode,
const tensor size,
const tensor stride,
const tensor input_offset = { 0,0,0,0 },
const padding output_padding = padding() 
)
inlinestatic

Constructs pooling primitive (computes input paddings to match output size).

Parameters
idThis primitive id.
inputInput primitive id.
modePooling mode.
strideDefines shift in input buffer between adjacent calculations of output values.
sizePooling kernel size.
output_sizeUser-defined output data size of the primitive (w/o padding).
Returns
Pooling primitive with specified settings.

Definition at line 116 of file pooling.hpp.


The documentation for this struct was generated from the following file: