Computer Senses User’s Frustration
Published: February 7, 2006
Wouldn’t it be great if your computer could recognize when you’re frustrated with it and adjust itself to calm you down? Emotion-sensing technology could someday allow a computer to do just that.
Computer scientist Christian Peter of the Fraunhofer Institute for Computer Graphics in Rostock, Germany, and his colleagues are working on a system that collects data about a person’s emotional state using sight, sound and touch technology.
The system then interprets the information and reacts accordingly.
For example, if a computer senses that its user is agitated, it might tone down the background color of the screen, turn down background music, enlarge or reduce graphics, adjust the flow of information being presented to the user or simply apologize.
“With humans, somebody who ignores the feelings of others is not liked as much as somebody who shows some sort of emotional feedback. Why should it be different with computers?” said Peter.
But sensing emotions from a person is not always an easy matter.
Current methods for collecting the data require researchers to wire users with electrodes and monitor their behavior in a laboratory setting.
Less-obtrusive means, such as using a video to monitor gestures or a recorder to analyze voice, allow the user to behave more naturally, but the data can break down if the person moves too far away.
Peter and his team are working on technology that unobtrusively senses a person’s emotions while they interact freely with a computer. Their latest prototype is a wireless electronic glove that measures heart rate, blood pressure and skin temperature.
Peter will be demonstrating the wireless glove at the CeBIT Exhibition in Hanover, Germany, in March.
“Fraunhofer has been on the leading of edge of innovating,” said computer scientist Rosalind Picard of the Massachusetts Institute of Technology.
Information gathered by the glove is transmitted wirelessly to a base unit, which stores it on a memory card or sends it to a computer database. Software written by Peter’s team analyzes the data and retrieves patterns that indicate certain emotions.
For example, if a person’s heart rate increases quickly and their skin temperature falls below a certain threshold, it may indicate that they are angry.
A different combination of other variables suggests the person is slightly surprised or very surprised.
The Fraunhofer team is also working on technology that will read facial features using an ordinary Web cam. The goal is to collect all of the emotion-indicating information in one database, analyze it in real time and program the computer to respond immediately.
If you enjoyed this good news Subscribe to Good News Blog
Share this
To share this simply copy and paste one of the below URL's: