Researchers have developed a PC-based system that can
recognise facial expressions in real time and deduce the person’s
mood.
The system was developed at Germany’s
Fraunhofer Institute for Integrated Circuits (IIS).
Possible applications include testing instinctive audience
response to
advertisements,
applications interfaces, information and driver alertness.
The system absorbs huge quantities of data containing images of
faces, mainly the contours of the face, the eyes, the eyebrows and
the nose and associates them with moods such as anger, happiness,
and puzzlement.
In operation, it compares 30,000 facial characteristics it
receives via a video link with the stored information.
Sophisticated algorithms localise human faces in the image,
differentiate between men and women, analyze their expressions and
deduce their mood.
Christian Küblbeck, project manager at the IIS, said the facial
analysis software operates in real time and can localise and
analyse many faces simultaneously.
“On a standard PC, the calculations are carried out so quickly
that mood changes can be tracked live,” said Küblbeck. “However, we
do not need to worry about an invasion of our privacy, as the
software analyzes the data on a purely statistical basis.”
Further trials to tackle challenges of biometrics >>
Face synthesis takes new line on recognition >>
Fraunhofer Institute for
Integrated Circuits >>
Comment on this article:
computer.weekly@rbi.co.uk