Researchers are working to develop a computerized lip-reading system that they say could be used to nab loose-lipped crooks.
University of East Anglia researchers in the United Kingdom are embarking on a US$766,000, 3-year project next month to create systems that convert videos of lip-motions into text. The project builds on speech reading systems already built by researchers at the school.
The university is joining forces with the Centre for Vision, Speech & Signal Processing at Surrey University, and England's Home Office Scientific Development Branch, which is keen on using the technology for crime fighting.
The researchers say lip-reading is a common practice for people who don't even realize they are doing it, but it is also an inexact science. What's more, the number of trained lip-readers is falling off as more people are taught sign language instead, the researchers say.
"We all lip read, for example in noisy situations like a bar or party, but even the performance of expert lip readers can be very poor," said Richard Harvey, senior lecturer at the university, in a statement. "It appears that the best lip-readers are the ones who learned to speak a language before they lost their hearing and who have been taught lip-reading intensively. It is a very desirable skill."
Among the possible applications for the technology could be installing a small camera on a mobile phone or car dashboard to support speech recognition (Others, such as IBM researchers, are also working on to change the way you drive via speech recognition technology.)