Active Perception Over Machine and Citizen Sensing

Today, many sensor networks and their applications employ a brute force approach to collecting and analyzing the huge volumes of sensor data currently generated. Such an approach often wastes valuable energy and computational resources by unnecessarily tasking sensors and generating observations of minimal use. People, on the other hand, have evolved sophisticated mechanisms to efficiently perceive their environment. This is accomplished through an ability to isolate the signal from the noise - by focusing attention and seeking out those observations containing useful information. With the rise of social networking technology, people are now empowered to share their observations and perceptions with the world. A satisfying integration of such knowledge with observations generated by machine sensors, however, remains elusive. In this talk, we describe and demonstrate a semantics driven active perception prototype - derived from cognitive theory - that may be used to more effectively integrate observations from both machine sensors and people in order to efficiently generate comprehensive and robust situation awareness