fast::SegmentationNetwork class
#include <FAST/Algorithms/NeuralNetwork/SegmentationNetwork.hpp>

Segmentation neural network process object.

Contents

This class is a convenience class for a neural network which performs segmentation by having 1 input image, and outputs 1 segmentation image. Internally it uses TensorToSegmentation. If you need multi-input or multi-output support, use NeuralNetwork with TensorToSegmentation instead.

Base classes

class NeuralNetwork
Neural network process object.

Public functions

auto create(std::string modelFilename, float scaleFactor, bool heatmapOutput, float threshold, bool hasBackgroundClass, float meanIntensity, float standardDeviationIntensity, bool resizeBackToOrigianlSize, std::vector<NeuralNetworkNode> inputNodes, std::vector<NeuralNetworkNode> outputNodes, std::string inferenceEngine, std::vector<std::string> customPlugins) -> std::shared_ptr<SegmentationNetwork>
Create instance.
auto create(std::string modelFilename, std::vector<NeuralNetworkNode> inputNodes, std::vector<NeuralNetworkNode> outputNodes, std::string inferenceEngine, std::vector<std::string> customPlugins) -> std::shared_ptr<SegmentationNetwork>
Create instance C++ friendly create with parameters that must be set before loading.
void setHeatmapOutput()
void setSegmentationOutput()
void setThreshold(float threshold)
auto getThreshold() const -> float
void setBackgroundClass(bool hasBackgroundClass)
void loadAttributes() virtual
void setResizeBackToOriginalSize(bool resize)
void setChannelsToIgnore(std::vector<int> channels)
Specify list of output channels to ignore.

Private functions

void execute() virtual

Function documentation

std::shared_ptr<SegmentationNetwork> fast::SegmentationNetwork::create(std::string modelFilename, float scaleFactor, bool heatmapOutput, float threshold, bool hasBackgroundClass, float meanIntensity, float standardDeviationIntensity, bool resizeBackToOrigianlSize, std::vector<NeuralNetworkNode> inputNodes, std::vector<NeuralNetworkNode> outputNodes, std::string inferenceEngine, std::vector<std::string> customPlugins)

Create instance.

Parameters
modelFilename path to model to load
scaleFactor A value which is multiplied with each pixel of input image before it is sent to the neural network. Use this to scale your pixels values. Default: 1.0
heatmapOutput If set to true, the output of this object will be a Tensor instead of a segmentation Image. Default: false
threshold The minimum value of the class confidence value to be accepted. Default: 0.5
hasBackgroundClass Whether the neural network has a channel 0 which represents the "background".
meanIntensity Mean intensity to subtract from each pixel of the input image
standardDeviationIntensity Standard deviation to divide each pixel of the input image by
resizeBackToOrigianlSize
inputNodes Specify names, and potentially shapes, of input nodes. Not necessary unless you only want to use certain inputs or specify the input shape manually.
outputNodes Specify names, and potentially shapes, of output nodes to use. Not necessary unless you only want to use certain outputs or specify the output shape manually.
inferenceEngine Specify which inference engine to use (TensorFlow, TensorRT, OpenVINO). By default, FAST will select the best inference engine available on your system.
customPlugins Specify path to any custom plugins/operators to load
Returns instance

std::shared_ptr<SegmentationNetwork> fast::SegmentationNetwork::create(std::string modelFilename, std::vector<NeuralNetworkNode> inputNodes, std::vector<NeuralNetworkNode> outputNodes, std::string inferenceEngine, std::vector<std::string> customPlugins)

Create instance C++ friendly create with parameters that must be set before loading.

Parameters
modelFilename Path to model to load
inputNodes Specify names, and potentially shapes, of input nodes. Not necessary unless you only want to use certain inputs or specify the input shape manually.
outputNodes Specify names, and potentially shapes, of output nodes to use. Not necessary unless you only want to use certain outputs or specify the output shape manually.
inferenceEngine Specify which inference engine to use (TensorFlow, TensorRT, OpenVINO). By default, FAST will select the best inference engine available on your system.
customPlugins Specify path to any custom plugins/operators to load
Returns instance

void fast::SegmentationNetwork::setThreshold(float threshold)

Parameters
threshold

Threshold to accept a channel X as being class X.

void fast::SegmentationNetwork::setBackgroundClass(bool hasBackgroundClass)

Parameters
hasBackgroundClass

Set whether channel 0 of segmentation tensor is the "background" class, thereby getting the label 0 in the resulting Segmentation.

void fast::SegmentationNetwork::setChannelsToIgnore(std::vector<int> channels)

Specify list of output channels to ignore.

Parameters
channels