Neural software listens for danger

The Smart Sensor Enabled Neural Threat Recognition and Identification (SENTRI) system from Safety Dynamics, Oak Brook, Ill., incorporates software developed by Dr. Ted Berger at the University of Southern California Center for Neural Engineering. The software mimics the way the human brain processes sounds in order to recognize, identify, and locate the source of suspicious noises.

“We´re now reaping the benefits of more than a decade of low-level funding,” explains Dr. Joel Davis, ONR program officer. “Dr. Berger was working on language, how humans understand words amid the noise of sounds they hear. He´s transferred that knowledge into engineering and mathematical concepts with other applications.”

Dr. Berger created mathematical models to mimic brain activity, and applied them to extract a signal from a noisy environment, which is what we do every time we understand the spoken word. He realized he could train software to distinguish sounds—such as a gun firing—from other noises, and even small-arms fire from that of heavy weapons.

The sound identification could be coupled with chemical or optical sensors, so that if exhaust fumes are detected at the same time as a weapons-like “bang,” the system would identify a truck backfiring rather than a weapon discharging. The Chicago and Los Angeles County police departments are testing the system to help fight crime in areas that are low on beat officers. Safety Dynamics received a Small Business Innovation Research Phase I award from the Navy in 2004 and is working on adapting its system for a lightweight mobile version that could be hand carried into the field.

From Office of Naval Research



The material in this press release comes from the originating research organization. Content may be edited for style and length. Want more? Sign up for our daily email.