Voice recognition

A technology that uses deep learning techniques to identify human voices, distinguish different human voices, and respond to human, spoken commands.

Spoken commands can range from telling a device to play a song to a text dictation or driving a car.

It uses so-called "voice biometrics", which anylyzes the frequency and pitch of a voice, to identify and separate two or more voices in a room. Take, for example, a Voice Assistant in a kitchen, where a family is cooking dinner. The device will have to tell the difference between the various people in the kitchen and whether they are talking to each other or giving the device a command ("Hey, play my favorite song!" or "Hey, how do you cook rice?"). In some cases, the device will have to determine whether a person is authorized to interact with it — where apps are aged-related or it's a case of gaining access to a house.

The technology can also used to authenticate a person's identity in banking, for instance. But this very same technology has given rise to phenomena such as "deepfake voice cloning" and "identity impersonation", which the Enterprise Europe Network (a European Union support network for small and medium-sized businesses) has described as a "growing threat." It says cloned voices created by advanced AI can be "completely indistinguishable from the real person by a human being." (za/fs)

 

Sources:

What is voice recognition? (ARM) https://www.arm.com/glossary/voice-recognition (accessed May 13, 2024)

Speech recognition (Britannica) https://www.britannica.com/technology/speech-recognition (accessed May 13, 2024)

Deepfake AI voice cloning detection against impersonation fraud (Enterprise Europe Network) https://een.ec.europa.eu/partnering-opportunities/deepfake-ai-voice-cloning-detection-against-impersonation-fraud (accessed May 13, 2024)

Read the rest of DW's AI Lexicon:

A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

We're keen to hear your feedback. Suggest an entry by sending us a comment. And let us know if you feel we have missed something, got it wrong, and tell us whether our AI Lexicon has helped you understand the technology better.

Written and edited by: Zulfikar Abbany (za), Fred Schwaller (fs)