Walking the Walk

ATLANTA -- Researchers at the Georgia Institute of Technology and elsewhere are developing technologies to recognize a person's gait.

Still in its infancy, the technology is gaining attention because of federal studies such as several Georgia Tech projects, funded by the federal Defense Advanced Research Projects Agency (DARPA).

One Georgia Tech study explores gait recognition by computer vision, and another takes an even more novel approach -- gait recognition with a radar system, similar to those police officers use to catch speeders.

The ultimate goal is to detect, classify and identify humans from as far away as 500 feet, day or night, and in all weather conditions. These capabilities would strengthen the protection of U.S. forces and facilities from terrorist attacks, according to DARPA officials.

Because gait-recognition technology is so new, researchers are assessing the uniqueness of gait and methods by which it can be evaluated.

With two years of experiments and analysis almost complete, researchers on both Georgia Tech projects are seeking continued funding for further studies. They still must address numerous technical issues; it will be at least five years before the technologies are commercialized, researchers said.

In the radar-system project, results from experiments, data analysis and algorithm design are promising, researchers said. The technique focuses on the gait cycle formed by the movements of a person's various body parts over time.

Researchers correctly identified 80 percent to 95 percent of individual subjects, with variances in that range during the three experiment days.

The next step is to build a more powerful radar system and subject it to lab and field testing. In experiments last year, subjects walked 50 feet away from the radar and then within 15 feet of it. Researchers are now building a radar system that can detect people from a distance of 500 feet or more.

In the study of gait recognition by computer vision, researchers identified subjects by focusing on "activity-specific static biometrics." These are static properties -- such as a person's leg length -- that can be measured from a single image.

Researchers also are developing statistical analysis tools that will allow them to use a small database with easily gathered data to predict how well a particular biometric, including gait recognition, will work on a larger population.

-- Georgia Institute of Technology

Raising Broadband Awareness

CONCORD, N.H. -- When the New Hampshire Division of Economic Development surveyed more than 500 businesses across the state about their awareness of high-speed Internet access availability, the results were surprising.

Just 25 percent of the businesses knew high-speed access was available in their area -- despite the fact that all cities in the state have at least 12 ISPs actively operating, said Stuart Arnett, director of the division.

"It was somewhat expected from the smaller companies, but that some of the larger companies still were a little confused [about the availability of broadband] was a surprise," Arnett said. "The bigger the company, the more likely they are to be aware of the accessibility of T1 [service] and related offerings."

Arnett, who also chairs the New Hampshire Telecommunications Development Advisory Board, believes it's the state's job to educate businesses about the benefits of high-speed Internet access as a productivity tool.

The board also found a vigorous local ISP industry, which also was a bit of a surprise.

"This local ISP industry, in addition to providing a roughly equivalent commodity, what they really excel at is the hand-holding," he said. "They're, often times, companies' IT departments. It's one of those things that, when you hear it, you say, 'Of course, that makes sense,' but it took us a while to realize it."