Voice biometrics technology, simply put, verifies the identity of a speaker. Somewhat confusingly, it’s also known as voice recognition, speaker recognition, voiceprinting, voice authentication and others. For the sake of clarity, we call it voice biometrics.
Speech Recognition Versus Voice Biometrics
Automated speech recognition (ASR) is a key aspect of technologies such as interactive voice response (IVR), in-vehicle entertainment and emergency services, mobile (e.g., smartphones) and computing. The technology identifies the words a speaker says.
Voice biometrics identifies specific speakers rather than the words they say. Each of our voices has distinguishing characteristics determined by our anatomy and behavioral speech patterns. The shape and size of our mouths and throats, as well as our language, pitch and speaking patterns (i.e., fast-talking versus slow-talking) all shape our voices.
Voice biometrics maps a speaker’s unique characteristics and then uses the map for later identification. A user provides one or more audio samples, which the system analyzes to create a unique voiceprint for the speaker. Whenever the user calls in, the software compares the speaker’s voiceprint to the voiceprint on file.
Applications of Voice Biometrics
Voice biometrics technology is key to speaker verification, where the technology is most common. Combined with an ID and a password or PIN, voice biometrics provides two-factor authentication. Researchers have even used voice biometrics to identify wolves in Yellowstone National Park by their howls.
- Telephone and internet transactions
- IVR-based banking systems
- Remote access to websites, and networks
- IVR-based health systems
- Audio signatures for digital documents
- Mobile workforces, including law enforcement and emergency services
- Online education
- Clinical research