DARPA Wants Supercharged Spy Cams
By Noah Shachtman
Wired
The Pentagon has a whole array of tools to snoop on its enemies. But those darn "Military Threats in the Global War on Terrorism (GWOT) use deceptive techniques to deny discovery by reconnaissance and surveillance sensors," a Defense Department document bemoans. So DARPA, the military's way-out research arm, has a slew of new programs to beef up its spy sensors.
Dynamic Multisensor Exploitation, or "DYME," aims to combine cameras, radar, and acoustic sensors to better find bad guys as they move through urban canyons, and along coastal waters.
The goal of "Building Labels for Urban Environments," or "BLUE" (you gotta love these acronyms), is to automatically label the structures seen in surveillance video. "Urban areas include many buildings of various types: hotels, stores, offices, apartments, fire stations, hospitals, restaurants, places of worship, etc." And many of those structures look a whole lot alike. "This poses a significant technical challenge for automated object recognition systems, which exploit patterns of visual features in still images."
How could BLUE tell the buildings apart?
Well, maybe spy drones could stare at a structure, and look at the people and vehicles moving around it for clues. "Successful technical approaches may transition to the system developed on the DARPA Urban Reasoning and Geospatial Exploitation (URGENT) program."
The Combat Video Analysis Engine (no acronym, alas) would use "computer vision, machine learning and probabilistic models to detect and recognize complex threats and suspicious activities without identification of specific individuals."
Dynamic Multisensor Exploitation, or "DYME," aims to combine cameras, radar, and acoustic sensors to better find bad guys as they move through urban canyons, and along coastal waters.
The goal of "Building Labels for Urban Environments," or "BLUE" (you gotta love these acronyms), is to automatically label the structures seen in surveillance video. "Urban areas include many buildings of various types: hotels, stores, offices, apartments, fire stations, hospitals, restaurants, places of worship, etc." And many of those structures look a whole lot alike. "This poses a significant technical challenge for automated object recognition systems, which exploit patterns of visual features in still images."
How could BLUE tell the buildings apart?
Well, maybe spy drones could stare at a structure, and look at the people and vehicles moving around it for clues. "Successful technical approaches may transition to the system developed on the DARPA Urban Reasoning and Geospatial Exploitation (URGENT) program."
The Combat Video Analysis Engine (no acronym, alas) would use "computer vision, machine learning and probabilistic models to detect and recognize complex threats and suspicious activities without identification of specific individuals."
No comments:
Post a Comment