Internetworkers: http://www.ibiblio.org/internetworkers/

["Alan MacHett" <machett AT ibiblio.org>]

http://www.technologyreview.com/articles/04/11/prototype71104.asp?p=1
//
A new audio surveillance system could help fight crime in the city and
protect kilometers of unmanned borders. Software developed by Ted Berger,
director of the University of Southern California Center for Neural
Engineering, can be trained to recognize and distinguish sounds that are
indicators of a security breach or a safety hazard, such as a gunshot or
the rattle of someone climbing a chain-link fence. The software is based
on mathematical models that mimic the way the brain interprets sound, but
it can distinguish between two similar sounds far more precisely than the
human ear. This fall, Oak Brook, IL-based Safety Dynamics plans to
implement the software in surveillance devices that monitor urban
activity. Mounted on streetlight poles, the devices will listen for
gunshots, then guide surveillance cameras toward the source of the sounds.
Berger says the technology can also be used for large-scale security; an
array of detectors placed along a deserted border, for example, could
listen for footfalls or whispers, painting a scene solely on the basis of
acoustic information. The detectors could then notify a central location
of any suspicious activity.
//