DARPA Begins Mind’s Eye Program
January 6, 2011
The Defense Advanced Research Projects Agency (DARPA) kicked off the Mind’s Eye Program earlier this week. The program’s focus is the development of smart cameras that can actually think for themselves. These systems will not just watch us but will also be able to analyze visual data-sets and assign significance to that data.
According to the DARPA news release, “Humans perform a wide range of visual tasks with ease, something no current artificial intelligence can do in a robust way. They [humans] have inherently strong spatial judgment and are able to learn new spatiotemporal concepts directly from the visual experience.”
Jobs like ‘persistent stare’, which involve camera equipped unmanned ground vehicles (UGVs), require hours of human analysis to make the footage captured useful by evaluating and differentiating operationally significant activities from trivial ones.
According to DARPA, the joint military community anticipates a notable increase in the role of UVG’s in support of future operations but they concede such a rollout would not be a force multiplier because they require humans to evaluate the data.
“A machine-based implementation of such abilities is broadly applicable to a wide range of applications, including ground surveillance.” DARPA announced.
In a system like Mind’s Eye the analysis can be done simultaneously utilizing an artificial visual intelligence which would result in the delivery of usable data in real-time.
DARPA has contracted with 12 research teams to develop the machine-based intelligence including JPL and Carnegie Mellon.
“These teams will develop a software subsystem suitable for employment on a camera for man-portable UGVs, integrating existing state of the art computer vision and AI while making novel contributions in visual event learning, new spatiotemporal representations, machine-generated envisionment, visual inspection and grounding of visual concepts.”
Three other teams have been tasked to develop system integration concepts: General Dynamics Robotic Systems, iRobot and Toyon Research Corporation. Working collaboratively, these teams will incorporate newly developed visual intelligence software into a camera package suitable for a man-portable UGV.