Artificial Intelligence in a Throughput Model by Rodgers Waymond;

Author:Rodgers, Waymond;
Language: eng
Format: epub
Publisher: Taylor & Francis Group
Published: 2020-05-15T00:00:00+00:00

Speaker (Voice) recognition

Voice-based digital assistants and telephone-based service portals are already using voice recognition to identify users and authenticate customers. Speaker or voice recognition diverges from speech recognition in that the former recognizes and identifies a speaker implementing voice biometrics and the latter scrutinizes what is being said. Voice biometrics comprises physical and behavioral characteristics. The physical features represent the shape of the vocal tract responsible for articulating and controlling speech production, whereas, the behavioral features includes pitch, cadence and tone, etc.

Voice biometric solutions consist of encoded frequencies or formats and produces a model “voice print” distinctive to an individual. This voice print is utilized for identification and authentication of the speaker. Moreover, in terms of collection, voice biometrics works by digitizing a profile of a person’s speech to produce a stored model voice print, or template. Biometric technology decreases each spoken word to fragments constituted of several dominant frequencies named formants. Each fragment has several tones that can be encapsulated in a digital format. The tones jointly identify the speaker’s unique voice print. Voice prints are stored in databases in a manner comparable to the warehousing of fingerprints or other biometric data.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.