Are voice recognition and voice biometrics the same thing?
Simply put, voice recognition identifies what words are being spoken by an individual, while voice biometrics identifies the person saying those words. Voice recognition, also known as speech recognition or speech-to-text technology, identifies spoken words and phrases and converts them into text-based content that can be read, understood and utilized by machines. A common use of speech recognition technology is enabling interactive voice response (IVR) systems to accept spoken commands, such as “I want to pay my bill,” rather than requiring the caller to press a number on their phone’s touch-tone pad to reach the billing department.
Voice biometrics applications use unique characteristics of a voice – such as cadence, accent, pitch and tone – that are influenced by physical traits, including the shape of a person’s larynx, lips, tongue and mouth, to identify the speaker. Voice biometrics solutions can either be “active,” dependent on the use of certain words or phrases in order to authenticate, or “passive,” with the ability to identify speakers during the normal flow of conversation. Both types of voice biometrics establish a baseline or “enrolled” voiceprint, then compare future conversations against the enrolled sample. Subsequent conversations that match the enrolled voiceprint allow speakers to be authenticated in the application, reducing or eliminating other, often time-consuming forms of caller verification. And conversations that don’t match the enrolled voiceprint help companies identify and avoid potentially fraudulent interactions.
Regardless of whether the active or passive method is used, the voiceprints are stored as mathematical representations of the voice features. They are not a recording and cannot be played, listened to, replicated, reverse-engineered or used outside of the system in which they were created, providing a secure method for organizations to store authentication information.