The result of a three-year study by researchers at the U.K. University of East Anglia discovered how lip-reading can be converted into text. The project is a joint undertaking between those scientists and the Centre for Vision, Speech, and Signal Processing at Surrey University. It builds on lip-reading work already done at UEA, those folks hope to collect thousands of hours of lip movements and feed them into a wide-ranging database, with the purpose being the ability to identify threatening phrases recorded on security cameras.
The scientists hope to even get this technology installed in mobile phones and vehicle dashboards, but we wonder about the efficacy of that plan. Having a security camera that can read lips and translate it into text seems much more plausible.
The technology would have to be sophisticated enough to identify the language that people are speaking, of course. And our take on that is this machine could spur people to take up all sorts of other languages, like Sanskrit and Hebrew, in order to avoid having security monitors hear them discuss what they’re going to have for dinner.