OpenCVForUnity.MlModule.EM Class Reference

The class implements the Expectation Maximization algorithm. More...

Public Member Functions
int	getClustersNumber ()
int	getCovarianceMatrixType ()
void	getCovs (List< Mat > covs)
	Returns covariation matrices.
Mat	getMeans ()
	Returns the cluster centers (means of the Gaussian mixture)
TermCriteria	getTermCriteria ()
double double double epsilon	getTermCriteriaAsValueTuple ()
Vec3d	getTermCriteriaAsVec3d ()
Mat	getWeights ()
	Returns weights of the mixtures.
override float	predict (Mat samples)
	Returns posterior probabilities for the provided samples.
override float	predict (Mat samples, Mat results)
	Returns posterior probabilities for the provided samples.
override float	predict (Mat samples, Mat results, int flags)
	Returns posterior probabilities for the provided samples.
double[]	predict2 (Mat sample, Mat probs)
	Returns a likelihood logarithm value and an index of the most probable mixture component for the given sample.
void	setClustersNumber (int val)
void	setCovarianceMatrixType (int val)
void	setTermCriteria (in Vec3d val)
void	setTermCriteria (in(double type, double maxCount, double epsilon) val)
void	setTermCriteria (TermCriteria val)
bool	trainE (Mat samples, Mat means0)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainE (Mat samples, Mat means0, Mat covs0)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainE (Mat samples, Mat means0, Mat covs0, Mat weights0)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainE (Mat samples, Mat means0, Mat covs0, Mat weights0, Mat logLikelihoods)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainE (Mat samples, Mat means0, Mat covs0, Mat weights0, Mat logLikelihoods, Mat labels)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainE (Mat samples, Mat means0, Mat covs0, Mat weights0, Mat logLikelihoods, Mat labels, Mat probs)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainEM (Mat samples)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainEM (Mat samples, Mat logLikelihoods)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainEM (Mat samples, Mat logLikelihoods, Mat labels)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainEM (Mat samples, Mat logLikelihoods, Mat labels, Mat probs)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainM (Mat samples, Mat probs0)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainM (Mat samples, Mat probs0, Mat logLikelihoods)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainM (Mat samples, Mat probs0, Mat logLikelihoods, Mat labels)
	Estimate the Gaussian mixture parameters from a samples set.
bool	trainM (Mat samples, Mat probs0, Mat logLikelihoods, Mat labels, Mat probs)
	Estimate the Gaussian mixture parameters from a samples set.
Public Member Functions inherited from OpenCVForUnity.MlModule.StatModel
float	calcError (TrainData data, bool test, Mat resp)
	Computes error on the training or test dataset.
override bool	empty ()
	Returns true if the Algorithm is empty (e.g. in the very beginning or after unsuccessful read.
int	getVarCount ()
	Returns the number of variables in training samples.
bool	isClassifier ()
	Returns true if the model is classifier.
bool	isTrained ()
	Returns true if the model is trained.
bool	train (Mat samples, int layout, Mat responses)
	Trains the statistical model.
bool	train (TrainData trainData)
	Trains the statistical model.
bool	train (TrainData trainData, int flags)
	Trains the statistical model.
Public Member Functions inherited from OpenCVForUnity.CoreModule.Algorithm
virtual void	clear ()
	Clears the algorithm state.
virtual string	getDefaultName ()
IntPtr	getNativeObjAddr ()
void	save (string filename)
Public Member Functions inherited from OpenCVForUnity.DisposableObject
void	Dispose ()
void	ThrowIfDisposed ()

Static Public Member Functions
static new EM	__fromPtr__ (IntPtr addr)
static EM	create ()
static EM	load (string filepath)
	Loads and creates a serialized EM from a file.
static EM	load (string filepath, string nodeName)
	Loads and creates a serialized EM from a file.
Static Public Member Functions inherited from OpenCVForUnity.MlModule.StatModel
static new StatModel	__fromPtr__ (IntPtr addr)
Static Public Member Functions inherited from OpenCVForUnity.CoreModule.Algorithm
static Algorithm	__fromPtr__ (IntPtr addr)
Static Public Member Functions inherited from OpenCVForUnity.DisposableObject
static IntPtr	ThrowIfNullIntPtr (IntPtr ptr)

Public Attributes
double double	maxCount
double	type

Static Public Attributes
const int	COV_MAT_DEFAULT = COV_MAT_DIAGONAL
const int	COV_MAT_DIAGONAL = 1
const int	COV_MAT_GENERIC = 2
const int	COV_MAT_SPHERICAL = 0
const int	DEFAULT_MAX_ITERS = 100
const int	DEFAULT_NCLUSTERS = 5
const int	START_AUTO_STEP = 0
const int	START_E_STEP = 1
const int	START_M_STEP = 2
Static Public Attributes inherited from OpenCVForUnity.MlModule.StatModel
const int	COMPRESSED_INPUT = 2
const int	PREPROCESSED_INPUT = 4
const int	RAW_OUTPUT = 1
const int	UPDATE_MODEL = 1

Protected Member Functions
override void	Dispose (bool disposing)
Protected Member Functions inherited from OpenCVForUnity.MlModule.StatModel
override void	Dispose (bool disposing)
Protected Member Functions inherited from OpenCVForUnity.CoreModule.Algorithm
Protected Member Functions inherited from OpenCVForUnity.DisposableOpenCVObject
	DisposableOpenCVObject ()
	DisposableOpenCVObject (bool isEnabledDispose)
	DisposableOpenCVObject (IntPtr ptr)
	DisposableOpenCVObject (IntPtr ptr, bool isEnabledDispose)
Protected Member Functions inherited from OpenCVForUnity.DisposableObject
	DisposableObject ()
	DisposableObject (bool isEnabledDispose)

Additional Inherited Members
Package Functions inherited from OpenCVForUnity.MlModule.StatModel
Package Functions inherited from OpenCVForUnity.CoreModule.Algorithm
Package Attributes inherited from OpenCVForUnity.DisposableOpenCVObject
Properties inherited from OpenCVForUnity.DisposableObject
bool	IsDisposed [get, protected set]
bool	IsEnabledDispose [get, set]

Detailed Description

The class implements the Expectation Maximization algorithm.

See also

ml_intro_em

Member Function Documentation

__fromPtr__()

static new EM OpenCVForUnity.MlModule.EM.__fromPtr__ ( IntPtr addr )

static

create()

static EM OpenCVForUnity.MlModule.EM.create ( )

static

Creates empty EM model. The model should be trained then using StatModel::train(traindata, flags) method. Alternatively, you can use one of the EM.train* methods or load it from file using Algorithm::load<EM>(filename).

Dispose()

override void OpenCVForUnity.MlModule.EM.Dispose ( bool disposing )

protectedvirtual

Reimplemented from OpenCVForUnity.CoreModule.Algorithm.

getClustersNumber()

int OpenCVForUnity.MlModule.EM.getClustersNumber

(

)

See also

setClustersNumber

getCovarianceMatrixType()

int OpenCVForUnity.MlModule.EM.getCovarianceMatrixType

(

)

See also

setCovarianceMatrixType

getCovs()

void OpenCVForUnity.MlModule.EM.getCovs

(

List< Mat >

covs

)

Returns covariation matrices.

Returns vector of covariation matrices. Number of matrices is the number of gaussian mixtures, each matrix is a square floating-point matrix NxN, where N is the space dimensionality.

getMeans()

Mat OpenCVForUnity.MlModule.EM.getMeans

(

)

Returns the cluster centers (means of the Gaussian mixture)

Returns matrix with the number of rows equal to the number of mixtures and number of columns equal to the space dimensionality.

getTermCriteria()

TermCriteria OpenCVForUnity.MlModule.EM.getTermCriteria

(

)

See also

setTermCriteria

getTermCriteriaAsValueTuple()

double double double epsilon OpenCVForUnity.MlModule.EM.getTermCriteriaAsValueTuple

(

)

getTermCriteriaAsVec3d()

Vec3d OpenCVForUnity.MlModule.EM.getTermCriteriaAsVec3d

(

)

See also

setTermCriteria

getWeights()

Mat OpenCVForUnity.MlModule.EM.getWeights

(

)

Returns weights of the mixtures.

Returns vector with the number of elements equal to the number of mixtures.

load() [1/2]

static EM OpenCVForUnity.MlModule.EM.load ( string filepath )

static

Loads and creates a serialized EM from a file.

Use EM.save to serialize and store an EM to disk. Load the EM from this file again, by calling this function with the path to the file. Optionally specify the node for the file containing the classifier

Parameters

filepath	path to serialized EM
nodeName	name of node containing the classifier

load() [2/2]

static EM OpenCVForUnity.MlModule.EM.load ( string filepath, string nodeName )

static

Loads and creates a serialized EM from a file.

Use EM.save to serialize and store an EM to disk. Load the EM from this file again, by calling this function with the path to the file. Optionally specify the node for the file containing the classifier

Parameters

filepath	path to serialized EM
nodeName	name of node containing the classifier

predict() [1/3]

override float OpenCVForUnity.MlModule.EM.predict ( Mat samples )

virtual

Returns posterior probabilities for the provided samples.

Parameters

samples	The input samples, floating-point matrix
results	The optional output \( nSamples \times nClusters\) matrix of results. It contains posterior probabilities for each sample from the input
flags	This parameter will be ignored

Reimplemented from OpenCVForUnity.MlModule.StatModel.

predict() [2/3]

override float OpenCVForUnity.MlModule.EM.predict ( Mat samples, Mat results )

virtual

Returns posterior probabilities for the provided samples.

Parameters

samples	The input samples, floating-point matrix
results	The optional output \( nSamples \times nClusters\) matrix of results. It contains posterior probabilities for each sample from the input
flags	This parameter will be ignored

Reimplemented from OpenCVForUnity.MlModule.StatModel.

predict() [3/3]

override float OpenCVForUnity.MlModule.EM.predict ( Mat samples, Mat results, int flags )

virtual

Returns posterior probabilities for the provided samples.

Parameters

samples	The input samples, floating-point matrix
results	The optional output \( nSamples \times nClusters\) matrix of results. It contains posterior probabilities for each sample from the input
flags	This parameter will be ignored

Reimplemented from OpenCVForUnity.MlModule.StatModel.

predict2()

double[] OpenCVForUnity.MlModule.EM.predict2	(	Mat	sample,
		Mat	probs )

Returns a likelihood logarithm value and an index of the most probable mixture component for the given sample.

Parameters

sample	A sample for classification. It should be a one-channel matrix of \(1 \times dims\) or \(dims \times 1\) size.
probs	Optional output matrix that contains posterior probabilities of each component given the sample. It has \(1 \times nclusters\) size and CV_64FC1 type.

The method returns a two-element double vector. Zero element is a likelihood logarithm value for the sample. First element is an index of the most probable mixture component for the given sample.

setClustersNumber()

void OpenCVForUnity.MlModule.EM.setClustersNumber

(

int

val

)

See also

getClustersNumber

setCovarianceMatrixType()

void OpenCVForUnity.MlModule.EM.setCovarianceMatrixType

(

int

val

)

See also

getCovarianceMatrixType

setTermCriteria() [1/3]

void OpenCVForUnity.MlModule.EM.setTermCriteria

(

in Vec3d

val

)

See also

getTermCriteria

setTermCriteria() [2/3]

void OpenCVForUnity.MlModule.EM.setTermCriteria

(

in(double type, double maxCount, double epsilon)

val

)

See also

getTermCriteria

setTermCriteria() [3/3]

void OpenCVForUnity.MlModule.EM.setTermCriteria

(

TermCriteria

val

)

See also

getTermCriteria

trainE() [1/6]

bool OpenCVForUnity.MlModule.EM.trainE	(	Mat	samples,
		Mat	means0 )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. You need to provide initial means \(a_k\) of mixture components. Optionally you can pass initial weights \(\pi_k\) and covariance matrices \(S_k\) of mixture components.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
means0	Initial means \(a_k\) of mixture components. It is a one-channel matrix of \(nclusters \times dims\) size. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
covs0	The vector of initial covariance matrices \(S_k\) of mixture components. Each of covariance matrices is a one-channel matrix of \(dims \times dims\) size. If the matrices do not have CV_64F type they will be converted to the inner matrices of such type for the further computing.
weights0	Initial weights \(\pi_k\) of mixture components. It should be a one-channel floating-point matrix with \(1 \times nclusters\) or \(nclusters \times 1\) size.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainE() [2/6]

bool OpenCVForUnity.MlModule.EM.trainE	(	Mat	samples,
		Mat	means0,
		Mat	covs0 )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. You need to provide initial means \(a_k\) of mixture components. Optionally you can pass initial weights \(\pi_k\) and covariance matrices \(S_k\) of mixture components.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
means0	Initial means \(a_k\) of mixture components. It is a one-channel matrix of \(nclusters \times dims\) size. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
covs0	The vector of initial covariance matrices \(S_k\) of mixture components. Each of covariance matrices is a one-channel matrix of \(dims \times dims\) size. If the matrices do not have CV_64F type they will be converted to the inner matrices of such type for the further computing.
weights0	Initial weights \(\pi_k\) of mixture components. It should be a one-channel floating-point matrix with \(1 \times nclusters\) or \(nclusters \times 1\) size.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainE() [3/6]

bool OpenCVForUnity.MlModule.EM.trainE	(	Mat	samples,
		Mat	means0,
		Mat	covs0,
		Mat	weights0 )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. You need to provide initial means \(a_k\) of mixture components. Optionally you can pass initial weights \(\pi_k\) and covariance matrices \(S_k\) of mixture components.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
means0	Initial means \(a_k\) of mixture components. It is a one-channel matrix of \(nclusters \times dims\) size. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
covs0	The vector of initial covariance matrices \(S_k\) of mixture components. Each of covariance matrices is a one-channel matrix of \(dims \times dims\) size. If the matrices do not have CV_64F type they will be converted to the inner matrices of such type for the further computing.
weights0	Initial weights \(\pi_k\) of mixture components. It should be a one-channel floating-point matrix with \(1 \times nclusters\) or \(nclusters \times 1\) size.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainE() [4/6]

bool OpenCVForUnity.MlModule.EM.trainE	(	Mat	samples,
		Mat	means0,
		Mat	covs0,
		Mat	weights0,
		Mat	logLikelihoods )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. You need to provide initial means \(a_k\) of mixture components. Optionally you can pass initial weights \(\pi_k\) and covariance matrices \(S_k\) of mixture components.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
means0	Initial means \(a_k\) of mixture components. It is a one-channel matrix of \(nclusters \times dims\) size. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
covs0	The vector of initial covariance matrices \(S_k\) of mixture components. Each of covariance matrices is a one-channel matrix of \(dims \times dims\) size. If the matrices do not have CV_64F type they will be converted to the inner matrices of such type for the further computing.
weights0	Initial weights \(\pi_k\) of mixture components. It should be a one-channel floating-point matrix with \(1 \times nclusters\) or \(nclusters \times 1\) size.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainE() [5/6]

bool OpenCVForUnity.MlModule.EM.trainE	(	Mat	samples,
		Mat	means0,
		Mat	covs0,
		Mat	weights0,
		Mat	logLikelihoods,
		Mat	labels )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. You need to provide initial means \(a_k\) of mixture components. Optionally you can pass initial weights \(\pi_k\) and covariance matrices \(S_k\) of mixture components.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
means0	Initial means \(a_k\) of mixture components. It is a one-channel matrix of \(nclusters \times dims\) size. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
covs0	The vector of initial covariance matrices \(S_k\) of mixture components. Each of covariance matrices is a one-channel matrix of \(dims \times dims\) size. If the matrices do not have CV_64F type they will be converted to the inner matrices of such type for the further computing.
weights0	Initial weights \(\pi_k\) of mixture components. It should be a one-channel floating-point matrix with \(1 \times nclusters\) or \(nclusters \times 1\) size.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainE() [6/6]

bool OpenCVForUnity.MlModule.EM.trainE	(	Mat	samples,
		Mat	means0,
		Mat	covs0,
		Mat	weights0,
		Mat	logLikelihoods,
		Mat	labels,
		Mat	probs )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. You need to provide initial means \(a_k\) of mixture components. Optionally you can pass initial weights \(\pi_k\) and covariance matrices \(S_k\) of mixture components.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
means0	Initial means \(a_k\) of mixture components. It is a one-channel matrix of \(nclusters \times dims\) size. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
covs0	The vector of initial covariance matrices \(S_k\) of mixture components. Each of covariance matrices is a one-channel matrix of \(dims \times dims\) size. If the matrices do not have CV_64F type they will be converted to the inner matrices of such type for the further computing.
weights0	Initial weights \(\pi_k\) of mixture components. It should be a one-channel floating-point matrix with \(1 \times nclusters\) or \(nclusters \times 1\) size.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainEM() [1/4]

bool OpenCVForUnity.MlModule.EM.trainEM

(

Mat

samples

)

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. Initial values of the model parameters will be estimated by the k-means algorithm.

Unlike many of the ML models, EM is an unsupervised learning algorithm and it does not take responses (class labels or function values) as input. Instead, it computes the Maximum Likelihood Estimate of the Gaussian mixture parameters from an input sample set, stores all the parameters inside the structure: \(p_{i,k}\) in probs, \(a_k\) in means , \(S_k\) in covs[k], \(\pi_k\) in weights , and optionally computes the output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample).

The trained model can be used further for prediction, just like any other classifier. The trained model is similar to the NormalBayesClassifier.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainEM() [2/4]

bool OpenCVForUnity.MlModule.EM.trainEM	(	Mat	samples,
		Mat	logLikelihoods )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. Initial values of the model parameters will be estimated by the k-means algorithm.

Unlike many of the ML models, EM is an unsupervised learning algorithm and it does not take responses (class labels or function values) as input. Instead, it computes the Maximum Likelihood Estimate of the Gaussian mixture parameters from an input sample set, stores all the parameters inside the structure: \(p_{i,k}\) in probs, \(a_k\) in means , \(S_k\) in covs[k], \(\pi_k\) in weights , and optionally computes the output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample).

The trained model can be used further for prediction, just like any other classifier. The trained model is similar to the NormalBayesClassifier.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainEM() [3/4]

bool OpenCVForUnity.MlModule.EM.trainEM	(	Mat	samples,
		Mat	logLikelihoods,
		Mat	labels )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. Initial values of the model parameters will be estimated by the k-means algorithm.

Unlike many of the ML models, EM is an unsupervised learning algorithm and it does not take responses (class labels or function values) as input. Instead, it computes the Maximum Likelihood Estimate of the Gaussian mixture parameters from an input sample set, stores all the parameters inside the structure: \(p_{i,k}\) in probs, \(a_k\) in means , \(S_k\) in covs[k], \(\pi_k\) in weights , and optionally computes the output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample).

The trained model can be used further for prediction, just like any other classifier. The trained model is similar to the NormalBayesClassifier.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainEM() [4/4]

bool OpenCVForUnity.MlModule.EM.trainEM	(	Mat	samples,
		Mat	logLikelihoods,
		Mat	labels,
		Mat	probs )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Expectation step. Initial values of the model parameters will be estimated by the k-means algorithm.

Unlike many of the ML models, EM is an unsupervised learning algorithm and it does not take responses (class labels or function values) as input. Instead, it computes the Maximum Likelihood Estimate of the Gaussian mixture parameters from an input sample set, stores all the parameters inside the structure: \(p_{i,k}\) in probs, \(a_k\) in means , \(S_k\) in covs[k], \(\pi_k\) in weights , and optionally computes the output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample).

The trained model can be used further for prediction, just like any other classifier. The trained model is similar to the NormalBayesClassifier.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainM() [1/4]

bool OpenCVForUnity.MlModule.EM.trainM	(	Mat	samples,
		Mat	probs0 )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Maximization step. You need to provide initial probabilities \(p_{i,k}\) to use this option.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
probs0	the probabilities
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainM() [2/4]

bool OpenCVForUnity.MlModule.EM.trainM	(	Mat	samples,
		Mat	probs0,
		Mat	logLikelihoods )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Maximization step. You need to provide initial probabilities \(p_{i,k}\) to use this option.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
probs0	the probabilities
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainM() [3/4]

bool OpenCVForUnity.MlModule.EM.trainM	(	Mat	samples,
		Mat	probs0,
		Mat	logLikelihoods,
		Mat	labels )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Maximization step. You need to provide initial probabilities \(p_{i,k}\) to use this option.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
probs0	the probabilities
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

trainM() [4/4]

bool OpenCVForUnity.MlModule.EM.trainM	(	Mat	samples,
		Mat	probs0,
		Mat	logLikelihoods,
		Mat	labels,
		Mat	probs )

Estimate the Gaussian mixture parameters from a samples set.

This variation starts with Maximization step. You need to provide initial probabilities \(p_{i,k}\) to use this option.

Parameters

samples	Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.
probs0	the probabilities
logLikelihoods	The optional output matrix that contains a likelihood logarithm value for each sample. It has \(nsamples \times 1\) size and CV_64FC1 type.
labels	The optional output "class label" for each sample: \(\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\) (indices of the most probable mixture component for each sample). It has \(nsamples \times 1\) size and CV_32SC1 type.
probs	The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \(nsamples \times nclusters\) size and CV_64FC1 type.

Member Data Documentation

COV_MAT_DEFAULT

const int OpenCVForUnity.MlModule.EM.COV_MAT_DEFAULT = COV_MAT_DIAGONAL

static

COV_MAT_DIAGONAL

const int OpenCVForUnity.MlModule.EM.COV_MAT_DIAGONAL = 1

static

COV_MAT_GENERIC

const int OpenCVForUnity.MlModule.EM.COV_MAT_GENERIC = 2

static

COV_MAT_SPHERICAL

const int OpenCVForUnity.MlModule.EM.COV_MAT_SPHERICAL = 0

static

DEFAULT_MAX_ITERS

const int OpenCVForUnity.MlModule.EM.DEFAULT_MAX_ITERS = 100

static

DEFAULT_NCLUSTERS

const int OpenCVForUnity.MlModule.EM.DEFAULT_NCLUSTERS = 5

static

maxCount

double double OpenCVForUnity.MlModule.EM.maxCount

START_AUTO_STEP

const int OpenCVForUnity.MlModule.EM.START_AUTO_STEP = 0

static

START_E_STEP

const int OpenCVForUnity.MlModule.EM.START_E_STEP = 1

static

START_M_STEP

const int OpenCVForUnity.MlModule.EM.START_M_STEP = 2

static

type

double OpenCVForUnity.MlModule.EM.type

See also

setTermCriteria

---

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

• OpenCVForUnity/Assets/OpenCVForUnity/org/opencv/ml/EM.cs
• OpenCVForUnity/Assets/OpenCVForUnity/org/opencv/ml/EM_Struct.cs
• OpenCVForUnity/Assets/OpenCVForUnity/org/opencv/ml/EM_ValueTuple.cs