Voiceprint models

A voiceprint model is a mathematical representation of the characteristics of the human voice. The model can be used in the process of identifying an individual. The vocal characteristics reflect differences between the anatomy of each individual, such as the size and shape of the throat and mouth.

Creation of voiceprint models

The voiceprint model, based on audio samples, is stored as a digital signature that is used later to identify the individual. The creation of a voiceprint model for a person is referred to as enrollment.

Audio used for enrollment

You can train a voiceprint model using recorded interactions and audio imported from external recording sources.