Science Scope

Coming soon: Robots in the sky that recognize and track you

Coming soon: Robots in the sky that recognize and track you

Posting in Science

Developing technologies will make drones more adept at recognizing the faces of specific individuals and understanding their intent -- even from afar.

Military research has been the source of a number of modern technologies, most notably the Internet.

But now, the Army just issued contracts to develop two technologies that don't seem as fun as, say, poking someone on Facebook.

The contracts, which Wired reports are for work on surveillance projects, could make drones more adept at targeting specific individuals.

One is to develop drones with strong facial recognition that prevents the drone from losing a face in a crowd. Others are for machines that can integrate intelligence data with information from an informant to determine your intent.

Part of a broader effort called TTL (for "Tagging, Tracking and Locating"), these new projects will support the Pentagon as it attempts to monitor enemies and insurgents in places like Afghanistan, where the strategy has switched from rebuilding societies to targeting specific individual bad actors.

Current technologies include using tiny transmitters that can use cellular, satellite or radio frequencies to report their whereabouts and lingering scents that mark targets with a vapor that can be tracked for hours. But they are inadequate because targets may discover their transmitters and remove them, and scents eventually dissipate.

A drone that recognizes you

Progeny Systems Corporation, which won one of the contracts, is developing a drone that can use photos to create a three-dimensional model of the target's face. As Wired says,

It’s not an easy trick to pull off — even with the proper lighting, and even with a willing subject. Building a model of someone on the run is harder. Constructing a model using the bobbing, weaving, flying, relatively low-resolution cameras on small unmanned aerial vehicles is tougher still.

The new technology, called the “Long Range, Non-cooperative, Biometric Tagging, Tracking and Location” system, could be revolutionary because it can overcome what is a current problem in tagging, tracking and locating work: targets are usually only visible occasionally in crowds or in sheltered positions.

Progeny's new project can take a poor-quality (50 pixel) photo of someone with any expression, in any pose and under any lighting and build a 3-D model of his/her face. After the face is initially entered into Progeny's system, it takes only another 15- or 20-pixel image to recognize him.

The technology is robust enough that it can tell identical twins apart, as evidenced by tests that researchers from Notre Dame and Michigan State Universities ran using images of faces at a “Twins Days” festival.

Though the software works better the closer the drone is, the facial information can be added to "soft biometric" information such as skin color, height, build, age and gender to track a person of interest from a distance too far to use facial recognition.

Drones that read your mind

Another technology, being developed by Charles River Analytics, analyzes human behavior to determine if someone has malicious intent. The technology, called Adversary Behavior Acquisition, Collection, Understanding, and Summarization (ABACUS), compiles behavioral data to determine if a subject has built up anger against the U.S. and might pose a threat.

Similarly, Modus Operandi, Inc. is developing a system that will use “probabilistic algorithms th[at] determine the likelihood of adversarial intent.” Its name is “Clear Heart,” which surely trades on the idea of transparency and does not imply what is to be found in these targets' hearts.

Photo: ijy/MorgueFile

via: Wired and Popular Science

Share this

Laura Shin

Features Editor

Laura Shin has been published in The New York Times, The Wall Street Journal and The Los Angeles Times, and is currently a contributor at Forbes. Previously, she worked at Newsweek, the New York Times, Wall Street Journal and LearnVest. She holds degrees from Stanford University and Columbia University's Graduate School of Journalism. Follow her on Twitter. Disclosure