write_samples_class_gmmT_write_samples_class_gmmWriteSamplesClassGmmWriteSamplesClassGmmwrite_samples_class_gmm (Operator)

Name

write_samples_class_gmmT_write_samples_class_gmmWriteSamplesClassGmmWriteSamplesClassGmmwrite_samples_class_gmm — Write the training data of a Gaussian Mixture Model to a file.

Signature

write_samples_class_gmm( : : GMMHandle, FileName : )

Description

write_samples_class_gmmwrite_samples_class_gmmWriteSamplesClassGmmWriteSamplesClassGmmwrite_samples_class_gmm writes the training samples stored in the Gaussian Mixture Model (GMM) GMMHandleGMMHandleGMMHandleGMMHandlegmmhandle to the file given by FileNameFileNameFileNamefileNamefile_name. write_samples_class_gmmwrite_samples_class_gmmWriteSamplesClassGmmWriteSamplesClassGmmwrite_samples_class_gmm can be used to build up a database of training samples, and hence to improve the performance of the GMM by training it with an extended data set (see train_class_gmmtrain_class_gmmTrainClassGmmTrainClassGmmtrain_class_gmm).

The file FileNameFileNameFileNamefileNamefile_name is overwritten by write_samples_class_gmmwrite_samples_class_gmmWriteSamplesClassGmmWriteSamplesClassGmmwrite_samples_class_gmm. Nevertheless, extending the database of training samples is easy because read_samples_class_gmmread_samples_class_gmmReadSamplesClassGmmReadSamplesClassGmmread_samples_class_gmm and add_sample_class_gmmadd_sample_class_gmmAddSampleClassGmmAddSampleClassGmmadd_sample_class_gmm add the training samples to the training samples that are already stored in memory with the GMM.

The created file can be read with read_samples_class_mlpread_samples_class_mlpReadSamplesClassMlpReadSamplesClassMlpread_samples_class_mlp if the classifier of a multilayer perceptron (MLP) should be used. The class of a training sample in the GMM corresponds to a component of the target vector in the MLP being 1.0.

Execution Information

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

Parameters

GMMHandleGMMHandleGMMHandleGMMHandlegmmhandle (input_control) class_gmm → (handle)

GMM handle.

FileNameFileNameFileNamefileNamefile_name (input_control) filename.write → (string)

File name.

Result

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