SampleFasterRCNN Class Reference

The SampleFasterRCNN class implements the FasterRCNN sample. More...

Collaboration diagram for SampleFasterRCNN:

Public Member Functions
	SampleFasterRCNN (const SampleFasterRCNNParams &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
void	constructNetwork (SampleUniquePtr< nvcaffeparser1::ICaffeParser > &parser, SampleUniquePtr< nvinfer1::IBuilder > &builder, SampleUniquePtr< nvinfer1::INetworkDefinition > &network, SampleUniquePtr< nvinfer1::IBuilderConfig > &config)
	Parses a Caffe model for FasterRCNN 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, handles post-processing of bounding boxes and verify results. More...
void	bboxTransformInvAndClip (const float *rois, const float *deltas, float *predBBoxes, const float *imInfo, const int N, const int nmsMaxOut, const int numCls)
	Performs inverse bounding box transform and clipping. More...
std::vector< int >	nonMaximumSuppression (std::vector< std::pair< float, int >> &scoreIndex, float *bbox, const int classNum, const int numClasses, const float nmsThreshold)
	Performs non maximum suppression on final bounding boxes. More...

Private Attributes
SampleFasterRCNNParams	mParams
	The parameters for the sample. More...
nvinfer1::Dims	mInputDims
	The dimensions of the input to the network. More...
std::vector< samplesCommon::PPM< kIMG_CHANNELS, kIMG_H, kIMG_W > >	mPPMs
	PPMs of test images. More...
std::shared_ptr< nvinfer1::ICudaEngine >	mEngine
	The TensorRT engine used to run the network. More...

Static Private Attributes
static const int	kIMG_CHANNELS = 3
static const int	kIMG_H = 375
static const int	kIMG_W = 500

Detailed Description

The SampleFasterRCNN class implements the FasterRCNN sample.

It creates the network using a caffe model

Member Typedef Documentation

SampleUniquePtr

template<typename T >

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

private

Constructor & Destructor Documentation

SampleFasterRCNN()

SampleFasterRCNN::SampleFasterRCNN ( const SampleFasterRCNNParams &  params )

inline

Member Function Documentation

build()

bool SampleFasterRCNN::build

(

)

Function builds the network engine.

Creates the network, configures the builder and creates the network engine.

This function creates the FasterRCNN network by parsing the caffe model and builds the engine that will be used to run FasterRCNN (mEngine)

Returns

Returns true if the engine was created successfully and false otherwise

Here is the call graph for this function:

infer()

bool SampleFasterRCNN::infer

(

)

Runs the TensorRT inference engine for this sample.

This function is the main execution function of the sample. It allocates the buffer, sets inputs and executes the engine.

Here is the call graph for this function:

teardown()

bool SampleFasterRCNN::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()

void SampleFasterRCNN::constructNetwork ( SampleUniquePtr< nvcaffeparser1::ICaffeParser > &  parser, SampleUniquePtr< nvinfer1::IBuilder > &  builder, SampleUniquePtr< nvinfer1::INetworkDefinition > &  network, SampleUniquePtr< nvinfer1::IBuilderConfig > &  config  )

private

Parses a Caffe model for FasterRCNN and creates a TensorRT network.

Uses a caffe parser to create the FasterRCNN network and marks the output layers.

Parameters

network	Pointer to the network that will be populated with the FasterRCNN network
builder	Pointer to the engine builder

Here is the call graph for this function:

Here is the caller graph for this function:

processInput()

bool SampleFasterRCNN::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:

Here is the caller graph for this function:

verifyOutput()

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

private

Filters output detections, handles post-processing of bounding boxes and verify results.

Filters output detections and handles post-processing of bounding boxes, verify result.

Returns

whether the detection output matches expectations

Here is the call graph for this function:

Here is the caller graph for this function:

bboxTransformInvAndClip()

void SampleFasterRCNN::bboxTransformInvAndClip ( const float *  rois, const float *  deltas, float *  predBBoxes, const float *  imInfo, const int  N, const int  nmsMaxOut, const int  numCls  )

private

Performs inverse bounding box transform and clipping.

Performs inverse bounding box transform.

Here is the caller graph for this function:

nonMaximumSuppression()

std::vector< int > SampleFasterRCNN::nonMaximumSuppression ( std::vector< std::pair< float, int >> &  scoreIndex, float *  bbox, const int  classNum, const int  numClasses, const float  nmsThreshold  )

private

Performs non maximum suppression on final bounding boxes.

Here is the caller graph for this function:

Member Data Documentation

mParams

SampleFasterRCNNParams SampleFasterRCNN::mParams

private

The parameters for the sample.

mInputDims

nvinfer1::Dims SampleFasterRCNN::mInputDims

private

The dimensions of the input to the network.

kIMG_CHANNELS

const int SampleFasterRCNN::kIMG_CHANNELS = 3

staticprivate

kIMG_H

const int SampleFasterRCNN::kIMG_H = 375

staticprivate

kIMG_W

const int SampleFasterRCNN::kIMG_W = 500

staticprivate

mPPMs

std::vector<samplesCommon::PPM<kIMG_CHANNELS, kIMG_H, kIMG_W> > SampleFasterRCNN::mPPMs

private

PPMs of test images.

mEngine

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

private

The TensorRT engine used to run the network.

---

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

• sampleFasterRCNN.cpp