The PC that’s good for your PMT

Researchers at Brunel University are developing a computer that will be able to sense the emotions of its user and offer them appropriate support.

The system will use sensing equipment to detect changes in the user’s pulse rate and breathing to provide an indication of any change in their mood. This will be coupled with a facial-recognition system that will plot changes in the user’s face to determine their emotions.

‘Social psychology studies have suggested that people express emotions to computers in the same way they do to other humans,’ said Dr Kate Hone of the university’s Department of Information Systems and Computing (DISC). ‘However, computers are missing out on the ability to aid us fully as they cannot currently understand this method of communication and interpret emotional signals.’

The team will monitor how people express their mood when they know they are working with an emotion-detecting machine. The project will pinpoint the periods when emotion-recognition technology will be most welcome by the user and aims to revise the traditional approach to technical support by creating a more intuitive and user-friendly system.

Applications already identified include intelligent tutors that change the content or pace of a computer tutorial based on the levels of interest or confusion registering on the face of the user and a CD player that selects music based on the user’s mood.

By integrating the system within the help function Brunel’s researchers believe that the computer will be able to respond to user frustration by attempting to solve a problem in a different way or offering advice. They also believe that the mood-recognition function could be used to create games or interactive movies.

The three-year project is sponsored by the Engineering and Physical Sciences Research Council (EPSRC), and will initially research integrating facial-recognition systems within a PC, as well as identifying variations in expression between different users experiencing the same emotions.

On the web