SampleUffFasterRcnn Class Reference

Collaboration diagram for SampleUffFasterRcnn:

Public Member Functions
	SampleUffFasterRcnn (const SampleUffFasterRcnnParams &params)
bool	build ()
	Function builds the network engine. More...
bool	infer ()
	Runs the TensorRT inference engine for this sample. More...
bool	teardown ()
	Cleans up any state created in the sample class. More...

Private Types
template<typename T >
using	SampleUniquePtr = std::unique_ptr< T, samplesCommon::InferDeleter >

Private Member Functions
bool	constructNetwork (SampleUniquePtr< nvinfer1::IBuilder > &builder, SampleUniquePtr< nvinfer1::INetworkDefinition > &network, SampleUniquePtr< nvuffparser::IUffParser > &parser)
	Parses an UFF model for SSD and creates a TensorRT network. More...
bool	processInput (const samplesCommon::BufferManager &buffers)
	Reads the input and mean data, preprocesses, and stores the result in a managed buffer. More...
bool	verifyOutput (const samplesCommon::BufferManager &buffers)
	Filters output detections and verify results. More...
void	batch_inverse_transform_classifier (const float *roi_after_nms, int roi_num_per_img, const float *classifier_cls, const float *classifier_regr, std::vector< float > &pred_boxes, std::vector< int > &pred_cls_ids, std::vector< float > &pred_probs, std::vector< int > &box_num_per_img, int N)
	Helper function to do post-processing(apply delta to ROIs). More...
std::vector< int >	nms_classifier (std::vector< float > &boxes_per_cls, std::vector< float > &probs_per_cls, float NMS_OVERLAP_THRESHOLD, int NMS_MAX_BOXES)
	NMS helper function in post-processing. More...
void	visualize_boxes (int img_num, int class_num, std::vector< float > &pred_boxes, std::vector< float > &pred_probs, std::vector< int > &pred_cls_ids, std::vector< int > &box_num_per_img, std::vector< vPPM > &ppms)
	Helper function to dump bbox-overlayed images as PPM files. More...

Private Attributes
SampleUffFasterRcnnParams	mParams
	The parameters for the sample. More...
nvinfer1::Dims	mInputDims
	The dimensions of the input to the network. More...
std::shared_ptr< nvinfer1::ICudaEngine >	mEngine
	The TensorRT engine used to run the network. More...

Member Typedef Documentation

SampleUniquePtr

template<typename T >

using SampleUffFasterRcnn::SampleUniquePtr = std::unique_ptr<T, samplesCommon::InferDeleter>

private

Constructor & Destructor Documentation

SampleUffFasterRcnn()

SampleUffFasterRcnn::SampleUffFasterRcnn ( const SampleUffFasterRcnnParams &  params )

inline

Member Function Documentation

build()

bool SampleUffFasterRcnn::build

(

)

Function builds the network engine.

Here is the call graph for this function:

infer()

bool SampleUffFasterRcnn::infer

(

)

Runs the TensorRT inference engine for this sample.

Here is the call graph for this function:

teardown()

bool SampleUffFasterRcnn::teardown

(

)

Cleans up any state created in the sample class.

Clean up the libprotobuf files as the parsing is complete

Note

It is not safe to use any other part of the protocol buffers library after ShutdownProtobufLibrary() has been called.

Here is the call graph for this function:

constructNetwork()

bool SampleUffFasterRcnn::constructNetwork ( SampleUniquePtr< nvinfer1::IBuilder > &  builder, SampleUniquePtr< nvinfer1::INetworkDefinition > &  network, SampleUniquePtr< nvuffparser::IUffParser > &  parser  )

private

Parses an UFF model for SSD and creates a TensorRT network.

Here is the call graph for this function:

processInput()

bool SampleUffFasterRcnn::processInput ( const samplesCommon::BufferManager &  buffers )

private

Reads the input and mean data, preprocesses, and stores the result in a managed buffer.

Here is the call graph for this function:

verifyOutput()

bool SampleUffFasterRcnn::verifyOutput ( const samplesCommon::BufferManager &  buffers )

private

Filters output detections and verify results.

Here is the call graph for this function:

batch_inverse_transform_classifier()

void SampleUffFasterRcnn::batch_inverse_transform_classifier ( const float *  roi_after_nms, int  roi_num_per_img, const float *  classifier_cls, const float *  classifier_regr, std::vector< float > &  pred_boxes, std::vector< int > &  pred_cls_ids, std::vector< float > &  pred_probs, std::vector< int > &  box_num_per_img, int  N  )

private

Helper function to do post-processing(apply delta to ROIs).

Define the function to apply delta to ROIs.

nms_classifier()

std::vector< int > SampleUffFasterRcnn::nms_classifier ( std::vector< float > &  boxes_per_cls, std::vector< float > &  probs_per_cls, float  NMS_OVERLAP_THRESHOLD, int  NMS_MAX_BOXES  )

private

NMS helper function in post-processing.

NMS on CPU in post-processing of classifier outputs.

visualize_boxes()

void SampleUffFasterRcnn::visualize_boxes ( int  img_num, int  class_num, std::vector< float > &  pred_boxes, std::vector< float > &  pred_probs, std::vector< int > &  pred_cls_ids, std::vector< int > &  box_num_per_img, std::vector< vPPM > &  ppms  )

private

Helper function to dump bbox-overlayed images as PPM files.

Dump the detection results(bboxes) as PPM images, overlayed on original image.

Here is the call graph for this function:

Member Data Documentation

mParams

SampleUffFasterRcnnParams SampleUffFasterRcnn::mParams

private

The parameters for the sample.

mInputDims

nvinfer1::Dims SampleUffFasterRcnn::mInputDims

private

The dimensions of the input to the network.

mEngine

std::shared_ptr<nvinfer1::ICudaEngine> SampleUffFasterRcnn::mEngine

private

The TensorRT engine used to run the network.

---

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

• sampleUffFasterRCNN.cpp