Single-Class Gaussian Mixture Model. More...

#include <xmmGmmSingleClass.hpp>

Inheritance diagram for xmm::SingleClassGMM:

Collaboration diagram for xmm::SingleClassGMM:

Public Member Functions
	SingleClassGMM (std::shared_ptr< SharedParameters > p=NULL)
	Constructor. More...

	SingleClassGMM (SingleClassGMM const &src)
	Copy constructor. More...

	SingleClassGMM (std::shared_ptr< SharedParameters > p, Json::Value const &root)
	Copy constructor. More...

SingleClassGMM &	operator= (SingleClassGMM const &src)
	Assignment. More...

Performance
void	reset ()
	Resets the fitering process (recognition or regression) More...

double	filter (std::vector< float > const &observation)
	filters a incoming observation (performs recognition or regression) More...

Json I/O
Json::Value	toJson () const
	Write the object to a JSON Structure. More...

void	fromJson (Json::Value const &root)
	Read the object from a JSON Structure. More...

Training
bool	isTraining () const
	Checks if the model is training. More...

void	train (TrainingSet *trainingSet)
	Main training method based on the EM algorithm. More...

void	cancelTraining ()
	Cancels the training process : sets a flag so that the training stops at the next possible exit in the training process. More...

Python File I/O
void	writeFile (char *fileName) const
	write method for python wrapping ('write' keyword forbidden, name has to be different) More...

void	readFile (char *fileName)
	read method for python wrapping ('read' keyword forbidden, name has to be different) More...

std::string	__str__ () const
	"print" method for python => returns the results of write method More...

Public Attributes
ClassParameters< GMM >	parameters
	Convert to bimodal GMM in place. More...

ClassResults< GMM >	results
	Results of the filtering process (recognition & regression) More...

std::vector< GaussianDistribution >	components
	Vector of Gaussian Components. More...

std::vector< float >	mixture_coeffs
	Mixture Coefficients. More...

std::vector< double >	beta
	Beta probabilities: likelihood of each component. More...

std::string	label
	label associated with the given model More...

std::shared_ptr< SharedParameters >	shared_parameters
	Pointer to the shared parameters owned by a multi-class model. More...

EventGenerator< TrainingEvent >	training_events
	Generator for events monitoring the training process. More...

TrainingEvent	training_status
	Event containing information on the current status of the training process. More...

Protected Member Functions
void	allocate ()
	Allocate model parameters. More...

double	obsProb (const float *observation, int mixtureComponent=-1) const
	Observation probability. More...

double	obsProb_input (const float *observation_input, int mixtureComponent=-1) const
	Observation probability on the input modality. More...

double	obsProb_bimodal (const float observation_input, const float observation_output, int mixtureComponent=-1) const
	Observation probability for bimodal mode. More...

void	emAlgorithmInit (TrainingSet *trainingSet)
	Initialize the EM Training Algorithm. More...

double	emAlgorithmUpdate (TrainingSet *trainingSet)
	Update Function of the EM algorithm. More...

void	initParametersToDefault (std::vector< float > const &dataStddev)
	Initialize model parameters to default values. More...

void	initMeansWithKMeans (TrainingSet *trainingSet)
	Initialize the means of the Gaussian components with a Biased K-means. More...

void	initCovariances_fullyObserved (TrainingSet *trainingSet)
	Initialize the Covariances of the Gaussian components using a fully observed approximation. More...

void	normalizeMixtureCoeffs ()
	Normalize mixture coefficients. More...

void	addCovarianceOffset ()
	Add offset to the diagonal of the covariance matrices. More...

void	updateInverseCovariances ()
	Update inverse covariances of each Gaussian component. More...

double	likelihood (std::vector< float > const &observation, std::vector< float > const &observation_output=null_vector_float)
	Compute likelihood and estimate components probabilities. More...

void	regression (std::vector< float > const &observation_input)
	Compute Gaussian Mixture Regression. More...

void	updateResults ()
	update the content of the likelihood buffer and return average likelihood. More...

virtual void	emAlgorithmTerminate ()
	Terminate the training algorithm. More...

bool	emAlgorithmHasConverged (int step, double log_prob, double old_log_prob) const
	checks if the training has converged according to the object's EM stop criterion More...

bool	cancelTrainingIfRequested ()
	checks if a cancel request has been sent and accordingly cancels the training process More...

void	check_training () const
	Checks if the model is still training. More...

Protected Attributes
std::vector< double >	current_regularization
	vector containing the regularization values over each dimension More...

CircularBuffer< double >	likelihood_buffer_
	Likelihood buffer used for smoothing. More...

std::mutex	training_mutex_
	Mutex used in Concurrent Mode. More...

bool	is_training_
	defines if the model is being trained. More...

bool	cancel_training_
	defines if the model received a request to cancel training More...

Friends
template<typename SingleClassModel , typename ModelType >
class	Model

class	SingleClassHMM

class	HierarchicalHMM

Detailed Description

Single-Class Gaussian Mixture Model.

Multivariate Gaussian Mixture Model. Supports Bimodal data and Gaussian Mixture Regression. Can be either autonomous or a state of a HMM: defines observation probabilities for each state.

Constructor & Destructor Documentation

xmm::SingleClassGMM::SingleClassGMM ( std::shared_ptr< SharedParameters > p = NULL )

Constructor.

Parameters

p	pointer to a shared parameters object (owned by a Model)

xmm::SingleClassGMM::SingleClassGMM ( SingleClassGMM const & src )

Copy constructor.

Parameters

src	Source GMM

xmm::SingleClassGMM::SingleClassGMM	(	std::shared_ptr< SharedParameters >	p,
		Json::Value const &	root
	)

explicit

Copy constructor.

Parameters

p	pointer to a shared parameters object (owned by a Model)
root	Json structure

Member Function Documentation

std::string xmm::Writable::__str__ ( ) const

inlineinherited

"print" method for python => returns the results of write method

Warning: only defined if SWIGPYTHON is defined

void xmm::SingleClassGMM::addCovarianceOffset ( )

protected

Add offset to the diagonal of the covariance matrices.

Guarantees convergence through covariance matrix invertibility

void xmm::SingleClassGMM::allocate ( )

protectedvirtual

Allocate model parameters.

Implements xmm::SingleClassProbabilisticModel.

void xmm::SingleClassProbabilisticModel::cancelTraining ( )

inherited

Cancels the training process : sets a flag so that the training stops at the next possible exit in the training process.

Warning: the model is still training when this function returns. This function only requests the training process to cancel.

bool xmm::SingleClassProbabilisticModel::cancelTrainingIfRequested ( )

protectedinherited

checks if a cancel request has been sent and accordingly cancels the training process

Returns: true if the training has been canceled.

void xmm::SingleClassProbabilisticModel::check_training ( ) const

inlineprotectedinherited

Checks if the model is still training.

Exceptions

runtime_error if the model is training.

bool xmm::SingleClassProbabilisticModel::emAlgorithmHasConverged	(	int	step,
		double	log_prob,
		double	old_log_prob
	)		const

protectedinherited

checks if the training has converged according to the object's EM stop criterion

Parameters

step	index of the current step of the EM algorithm
log_prob	log-likelihood returned by the EM update
old_log_prob	log-likelihood returned by the EM update at the previous step

void xmm::SingleClassGMM::emAlgorithmInit ( TrainingSet * trainingSet )

protectedvirtual

Initialize the EM Training Algorithm.

Initializes the Gaussian Components from the first phrase of the Training Set

Implements xmm::SingleClassProbabilisticModel.

void xmm::SingleClassProbabilisticModel::emAlgorithmTerminate ( )

protectedvirtualinherited

Terminate the training algorithm.

Reimplemented in xmm::SingleClassHMM.

double xmm::SingleClassGMM::emAlgorithmUpdate ( TrainingSet * trainingSet )

protectedvirtual

Update Function of the EM algorithm.

Returns: likelihood of the data given the current parameters (E-step)

Implements xmm::SingleClassProbabilisticModel.

double xmm::SingleClassGMM::filter ( std::vector< float > const & observation )

virtual

filters a incoming observation (performs recognition or regression)

the results of the inference process are stored in the results attribute

Parameters

observation observation vector

Returns: likelihood of the observation

Implements xmm::SingleClassProbabilisticModel.

void xmm::SingleClassGMM::fromJson ( Json::Value const & root )

virtual

Read the object from a JSON Structure.

Parameters

root	JSON value containing the object's information

Exceptions

JsonException if the JSON value has a wrong format

Implements xmm::SingleClassProbabilisticModel.

void xmm::SingleClassGMM::initCovariances_fullyObserved ( TrainingSet * trainingSet )

protected

Initialize the Covariances of the Gaussian components using a fully observed approximation.

void xmm::SingleClassGMM::initMeansWithKMeans ( TrainingSet * trainingSet )

protected

Initialize the means of the Gaussian components with a Biased K-means.

void xmm::SingleClassGMM::initParametersToDefault ( std::vector< float > const & dataStddev )

protected

Initialize model parameters to default values.

Mixture coefficients are then equiprobable

bool xmm::SingleClassProbabilisticModel::isTraining ( ) const

inherited

Checks if the model is training.

Returns: true if the model is training

double xmm::SingleClassGMM::likelihood	(	std::vector< float > const &	observation,
		std::vector< float > const &	observation_output = `null_vector_float`
	)

protected

Compute likelihood and estimate components probabilities.

If the model is bimodal, the likelihood is computed only on the input modality, except if 'observation_output' is specified. Updates the likelihood buffer used to smooth likelihoods.

Parameters

observation	observation vector (full size for unimodal, input modality for bimodal)
observation_output	observation vector of the output modality

void xmm::SingleClassGMM::normalizeMixtureCoeffs ( )

protected

Normalize mixture coefficients.

double xmm::SingleClassGMM::obsProb	(	const float *	observation,
		int	mixtureComponent = `-1`
	)		const

protected

Observation probability.

Parameters

observation	observation vector (must be of size 'dimension')
mixtureComponent	index of the mixture component. if unspecified or negative, full mixture observation probability is computed

Returns: likelihood of the observation given the model

Exceptions

out_of_range	if the index of the Gaussian Mixture Component is out of bounds
runtime_error	if the Covariance Matrix is not invertible

double xmm::SingleClassGMM::obsProb_bimodal	(	const float *	observation_input,
		const float *	observation_output,
		int	mixtureComponent = `-1`
	)		const

protected

Observation probability for bimodal mode.

Parameters

observation_input	observation vector of the input modality (must be of size 'dimension_input')
observation_output	observation vector of the input output (must be of size 'dimension - dimension_input')
mixtureComponent	index of the mixture component. if unspecified or negative, full mixture observation probability is computed

Returns: likelihood of the observation of the input modality given the model

Exceptions

runtime_error	if the model is not bimodal
runtime_error	if the Covariance Matrix is not invertible

double xmm::SingleClassGMM::obsProb_input	(	const float *	observation_input,
		int	mixtureComponent = `-1`
	)		const

protected

Observation probability on the input modality.

Parameters

observation_input	observation vector of the input modality (must be of size 'dimension_input')
mixtureComponent	index of the mixture component. if unspecified or negative, full mixture observation probability is computed

Returns: likelihood of the observation of the input modality given the model

Exceptions

runtime_error	if the model is not bimodal
runtime_error	if the Covariance Matrix of the input modality is not invertible

xmm::SingleClassGMM & xmm::SingleClassGMM::operator= ( SingleClassGMM const & src )

Assignment.

Parameters

src	Source GMM

void xmm::Writable::readFile ( char * fileName )

inlineinherited

read method for python wrapping ('read' keyword forbidden, name has to be different)

Warning: only defined if SWIGPYTHON is defined

void xmm::SingleClassGMM::regression ( std::vector< float > const & observation_input )

protected

Compute Gaussian Mixture Regression.

Estimates the output modality using covariance-based regression weighted by components' likelihood

Warning: the function does not estimates the likelihoods, use 'likelihood' before performing the regression.

Parameters

observation_input observation vector of the input modality

void xmm::SingleClassGMM::reset ( )

virtual

Resets the fitering process (recognition or regression)

Reimplemented from xmm::SingleClassProbabilisticModel.

Json::Value xmm::SingleClassGMM::toJson ( ) const

virtual

Write the object to a JSON Structure.

Returns: Json value containing the object's information

Reimplemented from xmm::SingleClassProbabilisticModel.

void xmm::SingleClassProbabilisticModel::train ( TrainingSet * trainingSet )

inherited

Main training method based on the EM algorithm.

the method performs a loop over the pure virtual method emAlgorithmUpdate() until convergence. The emAlgorithmUpdate method computes both E and M steps of the EM algorithm.

Parameters

trainingSet Training Set to train the model.

See also: emAlgorithmUpdate

void xmm::SingleClassGMM::updateInverseCovariances ( )

protected

Update inverse covariances of each Gaussian component.

Exceptions

runtime_error if one of the covariance matrices is not invertible

void xmm::SingleClassGMM::updateResults ( )

protected

update the content of the likelihood buffer and return average likelihood.

The method also updates the cumulative log-likelihood computed over a window (cumulativeloglikelihood)

void xmm::Writable::writeFile ( char * fileName ) const

inlineinherited

write method for python wrapping ('write' keyword forbidden, name has to be different)

Warning: only defined if SWIGPYTHON is defined

Friends And Related Function Documentation

friend class HierarchicalHMM

friend

template<typename SingleClassModel , typename ModelType >

friend class Model

friend

friend class SingleClassHMM

friend

Member Data Documentation

std::vector<double> xmm::SingleClassGMM::beta

Beta probabilities: likelihood of each component.

bool xmm::SingleClassProbabilisticModel::cancel_training_

protectedinherited

defines if the model received a request to cancel training

std::vector<GaussianDistribution> xmm::SingleClassGMM::components

Vector of Gaussian Components.

std::vector<double> xmm::SingleClassGMM::current_regularization

protected

vector containing the regularization values over each dimension

computed by combining relative and absolute regularization with the training set

bool xmm::SingleClassProbabilisticModel::is_training_

protectedinherited

defines if the model is being trained.

std::string xmm::SingleClassProbabilisticModel::label

inherited

label associated with the given model

CircularBuffer<double> xmm::SingleClassProbabilisticModel::likelihood_buffer_

protectedinherited

Likelihood buffer used for smoothing.

std::vector<float> xmm::SingleClassGMM::mixture_coeffs

Mixture Coefficients.

ClassParameters<GMM> xmm::SingleClassGMM::parameters

Convert to bimodal GMM in place.

Parameters

dimension_input dimension of the input modality

Exceptions

runtime_error	if the model is already bimodal
out_of_range	if the requested input dimension is too large Convert to unimodal GMM in place
runtime_error	if the model is already unimodal extract a submodel with the given columns

Parameters

columns columns indices in the target order

Exceptions

runtime_error	if the model is training
out_of_range	if the number or indices of the requested columns exceeds the current dimension

Returns: a GMM from the current model considering only the target columns extract the submodel of the input modality

Exceptions

runtime_error if the model is training or if it is not bimodal

Returns: a unimodal GMM of the input modality from the current bimodal model extract the submodel of the output modality

Exceptions

runtime_error if the model is training or if it is not bimodal

Returns: a unimodal GMM of the output modality from the current bimodal model extract the model with reversed input and output modalities

Exceptions

runtime_error if the model is training or if it is not bimodal

Returns: a bimodal GMM that swaps the input and output modalities Model Parameters

ClassResults<GMM> xmm::SingleClassGMM::results

Results of the filtering process (recognition & regression)

std::shared_ptr<SharedParameters> xmm::SingleClassProbabilisticModel::shared_parameters

inherited

Pointer to the shared parameters owned by a multi-class model.

EventGenerator<TrainingEvent> xmm::SingleClassProbabilisticModel::training_events

inherited

Generator for events monitoring the training process.

std::mutex xmm::SingleClassProbabilisticModel::training_mutex_

protectedinherited

Mutex used in Concurrent Mode.

TrainingEvent xmm::SingleClassProbabilisticModel::training_status

inherited

Event containing information on the current status of the training process.

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

XMM - Probabilistic Models for Motion Recognition and Mapping

Public Member Functions

Public Attributes

Protected Member Functions

Protected Attributes

Friends

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation

Friends And Related Function Documentation

Member Data Documentation