Computers Trump Humans in Reading Facial Cues of Pain
the MPR take:
In what could be a first, a software program has been developed by researchers at the University of California, San Diego that can detect genuine or faked pain in individual faces - and is more accurate than human interpreters. In a study published in the journal Current Biology, people who were asked to correctly interpret whether a facial expression of pain was real or fake had only a 50% success rate (comparable to guessing); even with an hour-long training session, a second group had a 55% success rate. The computer program Computer Expression Recognition Toolbox (CERT) made correct assessment 85% of the time by detecting minute muscle movement variations. Future developments include using the program to detect pain intensity in children for proper treatment.
How well can computers interact with humans. Researchers at the University of California, San Diego, have written software that not only detected whether a person's face revealed genuine or faked pain, but did so far more accurately than human observers. “A particular success like this has been ...
READ FULL ARTICLE From The New York Times