read_samples_class_gmmT_read_samples_class_gmmReadSamplesClassGmmReadSamplesClassGmmread_samples_class_gmm (Operator)

Name

read_samples_class_gmmT_read_samples_class_gmmReadSamplesClassGmmReadSamplesClassGmmread_samples_class_gmm — Read the training data of a Gaussian Mixture Model from a file.

Signature

read_samples_class_gmm( : : GMMHandle, FileName : )

Description

read_samples_class_gmmread_samples_class_gmmReadSamplesClassGmmReadSamplesClassGmmReadSamplesClassGmmread_samples_class_gmm reads training samples from the file given by FileNameFileNameFileNameFileNamefileNamefile_name and adds them to the training samples that have already been stored in the Gaussian Mixture Model (GMM) given by GMMHandleGMMHandleGMMHandleGMMHandleGMMHandlegmmhandle. The GMM must be created with create_class_gmmcreate_class_gmmCreateClassGmmCreateClassGmmCreateClassGmmcreate_class_gmm before calling read_samples_class_gmmread_samples_class_gmmReadSamplesClassGmmReadSamplesClassGmmReadSamplesClassGmmread_samples_class_gmm. As described with train_class_gmmtrain_class_gmmTrainClassGmmTrainClassGmmTrainClassGmmtrain_class_gmm and write_samples_class_gmmwrite_samples_class_gmmWriteSamplesClassGmmWriteSamplesClassGmmWriteSamplesClassGmmwrite_samples_class_gmm, read_samples_class_gmmread_samples_class_gmmReadSamplesClassGmmReadSamplesClassGmmReadSamplesClassGmmread_samples_class_gmm, add_sample_class_gmmadd_sample_class_gmmAddSampleClassGmmAddSampleClassGmmAddSampleClassGmmadd_sample_class_gmm, and write_samples_class_gmmwrite_samples_class_gmmWriteSamplesClassGmmWriteSamplesClassGmmWriteSamplesClassGmmwrite_samples_class_gmm can be used to build up a database of training samples, and hence to improve the performance of the GMM by retraining the GMM with extended data sets.

It should be noted that the training samples must have the correct dimensionality. The feature vectors stored in FileNameFileNameFileNameFileNamefileNamefile_name must have the lengths NumDimNumDimNumDimNumDimnumDimnum_dim that was specified with create_class_gmmcreate_class_gmmCreateClassGmmCreateClassGmmCreateClassGmmcreate_class_gmm, and enough classes must have been created in create_class_gmmcreate_class_gmmCreateClassGmmCreateClassGmmCreateClassGmmcreate_class_gmm. If this is not the case, an error message is returned.

It is possible to read files of samples that were written with write_samples_class_svmwrite_samples_class_svmWriteSamplesClassSvmWriteSamplesClassSvmWriteSamplesClassSvmwrite_samples_class_svm or write_samples_class_mlpwrite_samples_class_mlpWriteSamplesClassMlpWriteSamplesClassMlpWriteSamplesClassMlpwrite_samples_class_mlp.

Execution Information

Multithreading type: reentrant (runs in parallel with non-exclusive operators).
Multithreading scope: global (may be called from any thread).
Processed without parallelization.

This operator modifies the state of the following input parameter:

GMMHandleGMMHandleGMMHandleGMMHandleGMMHandlegmmhandle

During execution of this operator, access to the value of this parameter must be synchronized if it is used across multiple threads.

Parameters

GMMHandleGMMHandleGMMHandleGMMHandleGMMHandlegmmhandle (input_control, state is modified) class_gmm → (handle)

GMM handle.

FileNameFileNameFileNameFileNamefileNamefile_name (input_control) filename.read → (string)

File name.

Result

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