Real-Time Speaker Verification with a Microphone Array

Published

November 27, 2006

Author(s)

G. Mei, R. Xu, R. Lao, C. Kwan, Vincent M. Stanford

Abstract

Real-time speaker verification, with speech acquired using the NIST Mk-III microphone array and an autodirective beamforming algorithm, is demonstrated. The software and hardware backbone of the demonstration is the NIST Smart Flow System and Mk-III Array, both developed by National Institute of Standards and Technology in support of multimodal research communities. A microphone array acquires speech signals; a steered response beamformer calculates the direction of arrival (DoA) of the dominant signal; and a speaker verification component determines whether the signal is speech from a specific privileged speaker. If so, a camera will slew to the DoA of the privileged speaker; but not other speakers, or other kinds of sound. Novel approaches were taken to the design of the basic components to obtain good realtime demonstration performance.

Proceedings Title

World Congress in Computer Science, Computer Engineering, and Applied Computing | | | World Academy of Science

Conference Dates

June 26-29, 2006

Conference Location

Las Vegas, NV, USA

Conference Title

2006 International Conference on Pervasive Systems and Computing

Pub Type

Conferences

Download Paper

Local Download

Keywords

phased array processing, nearfield beamforming, speaker identification, autodirective speech acquisition, steered response beamforming

Citation

Mei, G. , Xu, R. , Lao, R. , Kwan, C. and Stanford, V. (2006), Real-Time Speaker Verification with a Microphone Array, World Congress in Computer Science, Computer Engineering, and Applied Computing | | | World Academy of Science, Las Vegas, NV, USA, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=50816 (Accessed December 7, 2024)