News Release

Finding a face in a crowd

Peer-Reviewed Publication

Office of Naval Research

The hot new tool on the market to ward off future terrorist attacks is a technology nurtured from its very beginnings by the Office of Naval Research, with later additional funding from other government agencies. Today, FaceIt® is a facial recognition software engine that allows computers to rapidly find, detect and recognize a face in the crowd.

Before FaceIt was a commercial product, it was a research program supported by the Office of Naval Research at Rockefeller University in New York City. The Office of Naval Research’s neural computation program supported this research as a promising model of the human visual system.

The FaceIt face recognition system is based on a procedure called local feature analysis, which captures tiny components of the face, allowing a face to be recognized when only a portion of the face is in view.

Today, FaceIt is a product of Visionics, Inc., of Jersey City, N.J., that can find human faces anywhere in a field of view. It can track up to 10 faces simultaneously in a live video. It can follow that face as it moves through a crowd while also searching for matches against a database at a rate of 60 million/minute, per central processing unit (CPU).

“In the process of studying human vision, Dr. Joseph Atick created the most powerful face recognition system in existence,” said Dr. Thomas McKenna, Office of Naval Research program manager for Dr. Atick’s research.

The technology has applications for surveillance, information security, access control, identity fraud, gang tracking, banking and finding missing children. It is currently being evaluated for use in airport security and as a counter-terrorism tool. Last February Dr. Atick’s work was selected by Technology Review magazine as one of 10 emerging technologies that will have a profound impact on our lives. For more information on the research behind the FaceIt technology if you are working media, please call Audrey Haar, 703-696-2869, or email haara@onr.navy.mil

###

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.