Public Member Functions | Protected Attributes
InferenceEngine::CNNNetwork Class Reference

This class contains all the information about the Neural Network and the related binary information. More...

#include <ie_cnn_network.h>

Collaboration diagram for InferenceEngine::CNNNetwork:
Collaboration graph
[legend]

Public Member Functions

  CNNNetwork ()=default
  A default constructor.
 
  CNNNetwork (std::shared_ptr< ICNNNetwork > network)
  Allows helper class to manage lifetime of network object. More...
 
  CNNNetwork (const std::shared_ptr< ngraph::Function > &network)
  A constructor from ngraph::Function object. More...
 
  CNNNetwork (std::shared_ptr< ICNNNetReader > reader)
  A constructor from ICNNNetReader object. More...
 
virtual  ~CNNNetwork ()
  A destructor.
 
virtual Precision  getPrecision () const
  Wraps original method ICNNNetwork::getPrecision.
 
virtual OutputsDataMap  getOutputsInfo () const
  Wraps original method ICNNNetwork::getOutputsInfo.
 
virtual InputsDataMap  getInputsInfo () const
  Wraps original method ICNNNetwork::getInputsInfo.
 
size_t  layerCount () const
  Wraps original method ICNNNetwork::layerCount.
 
const std::string &  getName () const noexcept
  Wraps original method ICNNNetwork::getName.
 
virtual void  setBatchSize (const size_t size)
  Wraps original method ICNNNetwork::setBatchSize.
 
virtual size_t  getBatchSize () const
  Wraps original method ICNNNetwork::getBatchSize.
 
  operator ICNNNetwork & () const
  An overloaded operator & to get current network. More...
 
const std::shared_ptr< ngraph::Function >  getFunction () const noexcept
  Returns constant nGraph function. More...
 
void  addOutput (const std::string &layerName, size_t outputIndex=0)
  Wraps original method ICNNNetwork::addOutput.
 
CNNLayerPtr  getLayerByName (const char *layerName) const
  Wraps original method ICNNNetwork::getLayerByName.
 
details::CNNNetworkIterator  begin () const
  Begin layer iterator Order of layers is implementation specific, and can be changed in future.
 
details::CNNNetworkIterator  end () const
  End layer iterator.
 
size_t  size () const
  number of layers in network object More...
 
void  AddExtension (InferenceEngine::IShapeInferExtensionPtr extension)
  Registers extension within the plugin. More...
 
virtual ICNNNetwork::InputShapes  getInputShapes () const
 
  • Helper method to get collect all input shapes with names of corresponding Data objects
More...
 
virtual void  reshape (const ICNNNetwork::InputShapes &inputShapes)
  Run shape inference with new input shapes for the network. More...
 
void  serialize (const std::string &xmlPath, const std::string &binPath="") const
  Serialize network to IR and weights files. More...
 

Protected Attributes

std::shared_ptr< ICNNNetReader reader
  reader extra reference, might be nullptr
 
std::shared_ptr< ICNNNetwork network
  network extra interface, might be nullptr
 
ICNNNetwork actual = nullptr
  A pointer to the current network.
 
DataPtr  output
  A pointer to output data.
 

Detailed Description

This class contains all the information about the Neural Network and the related binary information.

Constructor & Destructor Documentation

§ CNNNetwork() [1/3]

InferenceEngine::CNNNetwork::CNNNetwork ( std::shared_ptr< ICNNNetwork network )
inlineexplicit

Allows helper class to manage lifetime of network object.

Parameters
network Pointer to the network object

§ CNNNetwork() [2/3]

InferenceEngine::CNNNetwork::CNNNetwork ( const std::shared_ptr< ngraph::Function > &  network )
explicit

A constructor from ngraph::Function object.

Parameters
network Pointer to the ngraph::Function object

§ CNNNetwork() [3/3]

InferenceEngine::CNNNetwork::CNNNetwork ( std::shared_ptr< ICNNNetReader reader )
inlineexplicit

A constructor from ICNNNetReader object.

Parameters
reader Pointer to the ICNNNetReader object

Member Function Documentation

§ AddExtension()

void InferenceEngine::CNNNetwork::AddExtension ( InferenceEngine::IShapeInferExtensionPtr  extension )
inline

Registers extension within the plugin.

Parameters
extension Pointer to already loaded reader extension with shape propagation implementations

§ getFunction()

const std::shared_ptr<ngraph::Function> InferenceEngine::CNNNetwork::getFunction ( ) const
inlinenoexcept

Returns constant nGraph function.

Returns
constant nGraph function

§ getInputShapes()

virtual ICNNNetwork::InputShapes InferenceEngine::CNNNetwork::getInputShapes ( ) const
inlinevirtual

  • Helper method to get collect all input shapes with names of corresponding Data objects
Returns
Map of pairs: input's name and its dimension.

§ operator ICNNNetwork &()

InferenceEngine::CNNNetwork::operator ICNNNetwork & ( ) const
inline

An overloaded operator & to get current network.

Returns
An instance of the current network

§ reshape()

virtual void InferenceEngine::CNNNetwork::reshape ( const ICNNNetwork::InputShapes inputShapes )
inlinevirtual

Run shape inference with new input shapes for the network.

Parameters
inputShapes - map of pairs: name of corresponding data and its dimension.

§ serialize()

void InferenceEngine::CNNNetwork::serialize ( const std::string &  xmlPath,
const std::string &  binPath = "" 
) const
inline

Serialize network to IR and weights files.

Parameters
xmlPath Path to output IR file.
binPath Path to output weights file. The parameter is skipped in case of executable graph info serialization.

§ size()

size_t InferenceEngine::CNNNetwork::size ( ) const
inline

number of layers in network object

Returns

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