A group of scientists at UCSD’s Computer Vision and Robotics Research Laboratory recently received a $600,000 award to continue developing technology for an automated system designed to fight terrorism by detecting and tracking faces in a crowd.
The funding comes from the Technical Support Working Group, an interagency that provides for the research, development and prototyping of new technology for the National Research and Development Program for Combating Terrorism. The organization coordinates interagency and international research and development requirements for fighting terrorism.
The grant will be geared toward “”solving outstanding research problems,”” said Mohan Trivedi, CVRR director and electrical and computer engineering professor. Current and ongoing research projects associated with other agencies will also benefit from the grant. The goal is to refine the technical aspects of identifying and tracking possible terrorists through the use of networks of “”smart”” cameras.
There are two main technical problems researchers are trying to solve. First, camera systems must be able to automatically recognize significant events within a wide area and without human aid. Second, the cameras must work in a variety of conditions and environments. This includes functioning in different settings such as airports, navy bases and aircraft carriers. It also means that they have to work with different lighting conditions during daytime, night and any type of weather.
“”The systems must be smart enough to adapt,”” Trivedi said. “”In general, these are extremely difficult problems to solve.””
Because the money is coming from Washington, D.C., he explained, research will focus on national security.
“”[The award] gives us a very important research opportunity,”” Trivedi said. “”On another level, it also makes our work highly connected to making the community safer.””
He noted that the funding provides the resources they need to improve upon projects on which they are currently working.
The project is one of nine winning proposals for public safety and surveillance dealing with video technology. TSWG received more than 12,500 research proposals from industry, academic, government and national labs responding to a special call for bids after the Sept. 11 attacks.
Proposals went through three layers of review, narrowing down the competition to 600 entrants, and then to a few dozen.
Two prominent projects that the CVRR lab has been working on over the last five years involve using computer systems and “”smart”” cameras to improve traffic and automobile safety.
Sponsored mainly by Caltrans, the first project uses computer systems to identify roadway situations such as accidents and slow traffic. This then makes it easier to reach ambulances for help as well as contact family members. Multi-camera systems called Distributed Interactive Video Arrays have already been developed to monitor traffic. DIVAS connect various kinds of cameras, among which are Omnivision panoramic cameras.
The second project uses affective analysis to interpret human activity and make distracted drivers more aware. Researchers have worked with car companies like Nissan, Volkswagen and Chrysler to develop cooperative cameras that would make cars give signals to prompt people to pay more attention while driving.
The CVRR Laboratory at UCSD, associated with the Jacobs School of Engineering, has been researching the field of sensor-based intelligent systems for eight years. The group of approximately 20 researchers involved in the CVRR lab is comprised of undergraduate and graduate students, faculty members and visiting scholars. TSWG is an agency overseen by the Department of State’s coordinator for counterterrorism and is managed by the Department of Defense.
