You are here

Audio and Visual Recognition for the Advancement of Disaster Response and Humanitarian Relief

Award Information
Agency: Department of Defense
Branch: Air Force
Contract: FA8649-21-P-0054
Agency Tracking Number: FX20D-TCSO1-0273
Amount: $149,489.00
Phase: Phase I
Program: STTR
Solicitation Topic Code: AFX20D-TCSO1
Solicitation Number: X20.D
Solicitation Year: 2020
Award Year: 2021
Award Start Date (Proposal Award Date): 2021-01-06
Award End Date (Contract End Date): 2021-07-06
Small Business Information
555 Rose Ave Unit 1
Venice, CA 90291-1111
United States
DUNS: 079665564
HUBZone Owned: No
Woman Owned: No
Socially and Economically Disadvantaged: No
Principal Investigator
 Shaun Moore
 (847) 275-9377
Business Contact
 Shaun Moore
Phone: (847) 275-9377
Research Institution
 Miller Puckette
9500 Gilman Dr.
La Jolla, CA 92093-5004
United States

 (858) 534-4823
 Nonprofit College or University

In disaster response situations it is often necessary to use eVTOL/UAM vehicles to search for and identify humans when debris makes affected areas inaccessible on foot. We propose an auditory assist mechanism to detect the presence of human vocal sounds in the presence of background noise paired with visual recognition capabilities.  In an initial phase, we will assess the ability of trained machine learning platforms to determine whether a human voice is present in a synthetic training set, consisting of ambient sound with or without voices added.  The second phase will incorporate the visual spectrum and training a machine learning platform to use audio cues paired with automated video analysis.  If successful, in the following phase we would seek to integrate this technology into a prototype vocal and computer vision subsystem for search-and-rescue applications.  We would leverage both the audio and visual inputs of the scene to understand if an individual or group of individuals needs rescue.  

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

US Flag An Official Website of the United States Government