mlpack::ann::TripletMarginLoss< InputDataType, OutputDataType > Class Template Reference

The Triplet Margin Loss performance function measures the network's performance according to the relative distance from the anchor input of the positive (truthy) and negative (falsy) inputs. More...

#include <triplet_margin_loss.hpp>

Public Member Functions
	TripletMarginLoss (const double margin=1.0)
	Create the TripletMarginLoss object. More...
template<typename PredictionType , typename TargetType >
PredictionType::elem_type	Forward (const PredictionType &prediction, const TargetType &target)
	Computes the Triplet Margin Loss function. More...
template<typename PredictionType , typename TargetType , typename LossType >
void	Backward (const PredictionType &prediction, const TargetType &target, LossType &loss)
	Ordinary feed backward pass of a neural network. More...
OutputDataType &	OutputParameter () const
	Get the output parameter.
OutputDataType &	OutputParameter ()
	Modify the output parameter.
double	Margin () const
	Get the value of margin.
double &	Margin ()
	Modify the value of margin.
template<typename Archive >
void	serialize (Archive &ar, const unsigned int)
	Serialize the layer.

Detailed Description

template<typename InputDataType = arma::mat, typename OutputDataType = arma::mat>
class mlpack::ann::TripletMarginLoss< InputDataType, OutputDataType >

The Triplet Margin Loss performance function measures the network's performance according to the relative distance from the anchor input of the positive (truthy) and negative (falsy) inputs.

The distance between two samples A and B is defined as square of L2 norm of A-B.

For more information, refer the following paper.

@article{Schroff2015,
  author  = {Florian Schroff, Dmitry Kalenichenko, James Philbin},
  title   = {FaceNet: A Unified Embedding for Face Recognition and Clustering},
  year    = {2015},
  url     = {https://arxiv.org/abs/1503.03832},
}

Template Parameters

InputDataType	Type of the input data (arma::colvec, arma::mat, arma::sp_mat or arma::cube).
OutputDataType	Type of the output data (arma::colvec, arma::mat, arma::sp_mat or arma::cube).

Constructor & Destructor Documentation

TripletMarginLoss()

template<typename InputDataType , typename OutputDataType >

mlpack::ann::TripletMarginLoss< InputDataType, OutputDataType >::TripletMarginLoss

(

const double

margin = 1.0

)

Create the TripletMarginLoss object.

Parameters

margin

The minimum value by which the distance between Anchor and Negative sample exceeds the distance between Anchor and Positive sample.

Member Function Documentation

Backward()

template<typename InputDataType , typename OutputDataType >

template<typename PredictionType , typename TargetType , typename LossType >

void mlpack::ann::TripletMarginLoss< InputDataType, OutputDataType >::Backward	(	const PredictionType &	prediction,
		const TargetType &	target,
		LossType &	loss
	)

Ordinary feed backward pass of a neural network.

Parameters

prediction	Concatenated anchor and positive sample.
target	The negative sample.
loss	The calculated error.

Forward()

template<typename InputDataType , typename OutputDataType >

template<typename PredictionType , typename TargetType >

PredictionType::elem_type mlpack::ann::TripletMarginLoss< InputDataType, OutputDataType >::Forward	(	const PredictionType &	prediction,
		const TargetType &	target
	)

Computes the Triplet Margin Loss function.

Parameters

prediction	Concatenated anchor and positive sample.
target	The negative sample.

---

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

• src/mlpack/methods/ann/loss_functions/triplet_margin_loss.hpp
• src/mlpack/methods/ann/loss_functions/triplet_margin_loss_impl.hpp