Voice Print Analyzer is developed for speech analysis and classification, which is capable of analyzing audio files for Speech/Non-Speech Detection, Language Identification And Speaker Identification.

The core parts of VPA executing this analysis are called "classifiers". Based on a client /server architecture, VPA provides a software environment for:

	Training
	Configuration
	Classification
	Production

After classifier training and configuration, VPA automatically produces classification results. Training and Test result data as well as configuration data are stored in the data resource and are recalled for production.

After starting the production, VPA reads its input directory and processes the input files in a hierarchical order. Each input file represents a task for VPA. VPA processes them in the order of priorities. For each task, VPA produces result files and additionally displays the results on the screen. No further user intervention is required, but the user can observe the production progress and status. During production, VPA selects one of the configured signal flows (cascades) for each task. The selection depends on information provided by an external control unit. The file then runs through a cascade of classifiers assigned to the selected signal flow. In the cascade, the file is finally analyzed. This concept allows VPA to analyze signal files from different source.

A processing server coordinates the classification process. It encapsulates classification functions in a way that they are externally available as services.