classify_class_gmmT_classify_class_gmmClassifyClassGmmClassifyClassGmm (Operator)

Name

classify_class_gmmT_classify_class_gmmClassifyClassGmmClassifyClassGmm — Calculate the class of a feature vector by a Gaussian Mixture Model.

Signature

classify_class_gmm( : : GMMHandle, Features, Num : ClassID, ClassProb, Density, KSigmaProb)

Herror T_classify_class_gmm(const Htuple GMMHandle, const Htuple Features, const Htuple Num, Htuple* ClassID, Htuple* ClassProb, Htuple* Density, Htuple* KSigmaProb)

void ClassifyClassGmm(const HTuple& GMMHandle, const HTuple& Features, const HTuple& Num, HTuple* ClassID, HTuple* ClassProb, HTuple* Density, HTuple* KSigmaProb)

HTuple HClassGmm::ClassifyClassGmm(const HTuple& Features, Hlong Num, HTuple* ClassProb, HTuple* Density, HTuple* KSigmaProb) const

static void HOperatorSet.ClassifyClassGmm(HTuple GMMHandle, HTuple features, HTuple num, out HTuple classID, out HTuple classProb, out HTuple density, out HTuple KSigmaProb)

HTuple HClassGmm.ClassifyClassGmm(HTuple features, int num, out HTuple classProb, out HTuple density, out HTuple KSigmaProb)

Description

classify_class_gmmclassify_class_gmmClassifyClassGmmClassifyClassGmmClassifyClassGmm computes the best NumNumNumNumnum classes of the feature vector FeaturesFeaturesFeaturesFeaturesfeatures with the Gaussian Mixture Model (GMM) GMMHandleGMMHandleGMMHandleGMMHandleGMMHandle and returns the classes in ClassIDClassIDClassIDClassIDclassID and the corresponding probabilities of the classes in ClassProbClassProbClassProbClassProbclassProb. Before calling classify_class_gmmclassify_class_gmmClassifyClassGmmClassifyClassGmmClassifyClassGmm, the GMM must be trained with train_class_gmmtrain_class_gmmTrainClassGmmTrainClassGmmTrainClassGmm.

classify_class_gmmclassify_class_gmmClassifyClassGmmClassifyClassGmmClassifyClassGmm corresponds to a call to evaluate_class_gmmevaluate_class_gmmEvaluateClassGmmEvaluateClassGmmEvaluateClassGmm and an additional step that extracts the best NumNumNumNumnum classes. As described with evaluate_class_gmmevaluate_class_gmmEvaluateClassGmmEvaluateClassGmmEvaluateClassGmm, the output values of the GMM can be interpreted as probabilities of the occurrence of the respective classes. However, here the posterior probability ClassProbClassProbClassProbClassProbclassProb is further normalized as ClassProbClassProbClassProbClassProbclassProb = p(i|x)/p(x) , where p(i|x) and p(x) are specified with evaluate_class_gmmevaluate_class_gmmEvaluateClassGmmEvaluateClassGmmEvaluateClassGmm. In most cases it should be sufficient to use NumNumNumNumnum = 1 in order to decide whether the probability of the best class is high enough. In some applications it may be interesting to also take the second best class into account (NumNumNumNumnum = 2), particularly if it can be expected that the classes show a significant degree of overlap.

DensityDensityDensityDensitydensity and KSigmaProbKSigmaProbKSigmaProbKSigmaProbKSigmaProb are explained with evaluate_class_gmmevaluate_class_gmmEvaluateClassGmmEvaluateClassGmmEvaluateClassGmm.

Execution Information

Parameters

GMMHandleGMMHandleGMMHandleGMMHandleGMMHandle (input_control)  class_gmm HClassGmm, HTupleHTupleHtuple (handle) (IntPtr) (HHandle) (handle)

GMM handle.

FeaturesFeaturesFeaturesFeaturesfeatures (input_control)  real-array HTupleHTupleHtuple (real) (double) (double) (double)

Feature vector.

NumNumNumNumnum (input_control)  integer HTupleHTupleHtuple (integer) (int / long) (Hlong) (Hlong)

Number of best classes to determine.

Default value: 1

Suggested values: 1, 2, 3, 4, 5

ClassIDClassIDClassIDClassIDclassID (output_control)  integer(-array) HTupleHTupleHtuple (integer) (int / long) (Hlong) (Hlong)

Result of classifying the feature vector with the GMM.

ClassProbClassProbClassProbClassProbclassProb (output_control)  real-array HTupleHTupleHtuple (real) (double) (double) (double)

A-posteriori probability of the classes.

DensityDensityDensityDensitydensity (output_control)  real-array HTupleHTupleHtuple (real) (double) (double) (double)

Probability density of the feature vector.

KSigmaProbKSigmaProbKSigmaProbKSigmaProbKSigmaProb (output_control)  real-array HTupleHTupleHtuple (real) (double) (double) (double)

Normalized k-sigma-probability for the feature vector.

Result

If the parameters are valid, the operator classify_class_gmmclassify_class_gmmClassifyClassGmmClassifyClassGmmClassifyClassGmm returns the value 2 (H_MSG_TRUE). If necessary an exception is raised.

Possible Predecessors

train_class_gmmtrain_class_gmmTrainClassGmmTrainClassGmmTrainClassGmm, read_class_gmmread_class_gmmReadClassGmmReadClassGmmReadClassGmm

Alternatives

evaluate_class_gmmevaluate_class_gmmEvaluateClassGmmEvaluateClassGmmEvaluateClassGmm

See also

create_class_gmmcreate_class_gmmCreateClassGmmCreateClassGmmCreateClassGmm

References

Christopher M. Bishop: “Neural Networks for Pattern Recognition”; Oxford University Press, Oxford; 1995.
Mario A.T. Figueiredo: “Unsupervised Learning of Finite Mixture Models”; IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 3; March 2002.

Module

Foundation