You are here

Acoustic Pattern Recognition for Security Breaching Noise Detection

Award Information
Agency: Department of Defense
Branch: Navy
Contract: N00014-06-M-0199
Agency Tracking Number: N064-036-0171
Amount: $69,984.00
Phase: Phase I
Program: STTR
Solicitation Topic Code: N06-T036
Solicitation Number: N/A
Timeline
Solicitation Year: 2006
Award Year: 2006
Award Start Date (Proposal Award Date): 2006-08-01
Award End Date (Contract End Date): 2007-05-31
Small Business Information
6300 Gateway Dr.
Cypress, CA 90630
United States
DUNS: 614108918
HUBZone Owned: No
Woman Owned: No
Socially and Economically Disadvantaged: No
Principal Investigator
 Marvin Velazquez
 Principal Investigator
 (714) 224-4410
 mvelazquez@sara.com
Business Contact
 Parviz Parhami
Title: Chief Executive Officer
Phone: (714) 224-4410
Email: pparhami@sara.com
Research Institution
 DUKE UNIV.
 Leslie M Collins
 
ECE Department Duke University
Durham, NC 27708
United States

 (919) 660-5260
 Nonprofit college or university
Abstract

A proposed acoustic pattern recognition system will be based around an array of acoustic sensors to allow localization of sound events and cancellation of background noise. The number of sensors could be adjusted to vary the coverage area depending on the site. The measured signals will provide the input to a sophisticated signal processing algorithm that will separate the acoustic signals into constituent sound sources in order to detect, characterize, localize, and identify security breaching sounds, all amongst a high background noise environment. Following source separation, an adaptive classifier will be utilized to discriminate between nuisance and background sounds and security breaching sounds. A novel attribute of this classifier is its feature extraction approach, which leverages the images produced by auditory models and thus tightly couples the classifier to the mechanisms potentially used by the biological systems we propose to mimic.BENEFITS: lotsDevelopment of this sensor capability will provide a major step in developing a sensor recognition system that can detect the sounds of break-on and gunfire. It will support a number of military applications besides the facility security mission, such as monitoring trail use and protecting troops from infiltrators and ambush. Commercially, the increased need for a border, plant, and perimeter security sensor that can reduce the need for extensive manpower would be welcomed.

* Information listed above is at the time of submission. *

US Flag An Official Website of the United States Government