Thursday, December 16, 2010

On gait recognition (unless you're a U.S. government employee)


I subscribe to several biometric feeds and mailing lists, and one of them mentioned some biometric research that is being conducted by the government of China.

[NOTE: IF YOU ARE AN EMPLOYEE OF THE U.S. GOVERNMENT, YOU MAY NEED TO STOP READING THIS. THE NEW YORK TIMES HAS QUOTED A DIRECTIVE FROM THE OFFICE OF MANAGEMENT AND BUDGET THAT READS, IN PART, "Classified information, whether or not already posted on public websites or disclosed to the media, remains classified, and must be treated as such by federal employees and contractors, until it is declassified by an appropriate U.S. Government authority."]

The feed linked to a Zeenews article that quoted from a Wikileaks cable marked as confidential (and available from the Guardian, which has not yet been classified as a terrorist organization).

The cable in question discussed work by the Chinese Academy of Science (CAS) Institute of Intelligent Machines (IIM) in Hefei.

View Larger Map

Specifically, the IIM was working on gait recognition. According to the cable, the gait of an individual is measured by weight, as well as "two-dimensional sheer forces."

Non-classified information from a University of Southampton paper provides related background. Here is the abstract:

This paper describes the development of a prototype floor sensor as a gait recognition system. This could eventually find deployment as a standalone system (eg. a burglar alarm system) or as part of a multimodal biometric system. The new sensor consists of 1536 individual sensors arranged in a 3m by 0.5m rectangular strip with an individual sensor area of 3 cm2. The sensor floor operates at a sample rate of 22 Hz. The sensor itself uses a simple design inspired by computer keyboards and is made from low cost, off the shelf materials. Application of the sensor floor to a small database of 15 individuals was performed. Three features were extracted : stride length, stride cadence, and time on toe to time on heel ratio. Two of these measures have been used in video based gait recognition while the third is new to this analysis. These features proved sufficient to achieve an 80% recognition rate.

For more information on gait recognition, see this page from Global Security Intelligence. It turns out that IIM (mascot: The Fighting Bots) is not the only research institute that is investigating this.

The technology is moving ahead at a rapid pace, however, with government-sponsored projects supporting research such as that going on at the Georgia Institute of Technology, MIT, the Lappeenranta University of Technology, and others [sic] academic institutions.

Lappeenranta, incidentally, is in Finland, but luckily for me, its web pages are in English. The university has a machine vision and pattern recognition laboratory, and while its web page does not explicitly mention gait recognition, another biometric method is discussed:

The Laboratory, lead by Professor Ville Kyrki, educates experts of intelligent computing and produces new machine vision and pattern recognition solutions. The goal is to build usefull applications for industry and citizens. For example, research is focused on face recognition for a biometric passport and machine vision systems for automatic paper and board printability tests. In medical image processing retina images are studied for the efficient healthcare of diabetes.

I hear, however, that the Laboratory's gait recognition studies hit a snag when they selected their first two subjects for study - Kiira Korpi and Laura Lepisto. I am about to print information that even Wikileaks doesn't have, and which may be the double-secret transcripts of LUT's gait recognition study:

The two subjects are on figure skates.

And that, my friends, is what passes for a joke on the Empoprise-BI business blog...
blog comments powered by Disqus