vart::PostProcessImplBase Class Reference

Abstract interface for post-processing implementation used by PostProcess. More...

#include <vart_postprocess_impl_base.hpp>

Public Member Functions
	PostProcessImplBase ()=default
	Default construction is allowed for derived implementations. More...
virtual	~PostProcessImplBase ()
	Virtual destructor to ensure proper cleanup of derived implementations. More...
virtual void	set_config (std::vector< TensorInfo > &info, uint32_t batch_size)=0
	set_config() - Set PostProcessInfo config data before start doing the post-process. More...
virtual std::vector< std::vector< std::shared_ptr< InferResult > > >	process (std::vector< int8_t * > data, uint32_t current_batch_size)=0
	process() - Process/parse tensors data from ML network output to create infer results. More...
virtual std::vector< std::vector< std::shared_ptr< InferResult > > >	process (std::vector< std::vector< std::shared_ptr< vart::Memory >>> tensor_memory, uint32_t current_batch_size)=0
	process() - Process/parse tensors data from ML network output to create infer results. More...

Detailed Description

Abstract interface for post-processing implementation used by PostProcess.

Note

Applications should use vart::PostProcess rather than this type directly, unless extending or integrating a new backend.

Constructor & Destructor Documentation

PostProcessImplBase()

vart::PostProcessImplBase::PostProcessImplBase ( )

default

Default construction is allowed for derived implementations.

~PostProcessImplBase()

virtual vart::PostProcessImplBase::~PostProcessImplBase ( )

inlinevirtual

Virtual destructor to ensure proper cleanup of derived implementations.

Member Function Documentation

process() [1/2]

virtual std::vector<std::vector<std::shared_ptr<InferResult> > > vart::PostProcessImplBase::process ( std::vector< int8_t * >  data, uint32_t  current_batch_size  )

pure virtual

process() - Process/parse tensors data from ML network output to create infer results.

Parameters

[in]	data	Array of tensors data. Each tensor will have the data for the entire batch of images.
[in]	current_batch_size	Number of inputs in the current batch.

Returns

Vector of inference result objects for every image in the batch.

process() [2/2]

virtual std::vector<std::vector<std::shared_ptr<InferResult> > > vart::PostProcessImplBase::process ( std::vector< std::vector< std::shared_ptr< vart::Memory >>>  tensor_memory, uint32_t  current_batch_size  )

pure virtual

process() - Process/parse tensors data from ML network output to create infer results.

Parameters

[in]	tensor_memory	Vector of vart::Memory pointer groups. Each outer entry corresponds to one batch element; each inner entry is the set of output-tensor buffers for that element.
[in]	current_batch_size	Number of inputs in the current batch.

Returns

Vector of inference result objects for every image in the batch.

set_config()

virtual void vart::PostProcessImplBase::set_config ( std::vector< TensorInfo > &  info, uint32_t  batch_size  )

pure virtual

set_config() - Set PostProcessInfo config data before start doing the post-process.

Parameters

[in]	info	TensorInfo to be set.
[in]	batch_size	Supported batch size.

Use this method to set batch size and tensor information required to parse/process the ML network output. Call this method before the first call to process() to ensure proper configuration of the post-processing logic.

---

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

• vart_postprocess_impl_base.hpp