Computer scientists in Germany and California have developed a way to capture and map facial expressions from one user to another. When one person frowns, for instance, that expression is transferred to a live video of the other user and mapped on to their face. By running the process with good-quality cameras and on several computers at once, latency is sub-30ms. The technology could find applications in modifying dubbed video to match new languages. Watch a video at http://tinyurl.com/nhzvu3f.
Analyst Comment
This seemed sufficiently ‘creepy’ for our unofficial Halloween issue. (TA)
