BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Detecting engagement and positive emotions from facial expressions
  - Hamed Monkaresi
DTSTART:20130530T101500Z
DTEND:20130530T111500Z
UID:TALK45583@talks.cam.ac.uk
CONTACT:Henrik Lieng
DESCRIPTION:Engagement is often associated with better learning and higher
 \nproductivity\, but it is also linked with having positive experiences. A
  system\nthat can detect and support engaged states would also be contribu
 ting to the\nusers’ well-being. Psychologists have developed methods for
  measuring different\nforms of engagement\, but they require time consumin
 g manual coding and have not\nyet exploited new affective computing techni
 ques for its automation. In this\npaper we explore how the webcam\, and th
 e more recent Kinect depth camera can be\nused to collect and interpret ob
 servational and self-reported data. The study\nintroduces computer vision 
 techniques to make inferences about engagement and\nemotional states\, whi
 le students (N=23) engage in a structured writing activity\n(draft-feedbac
 k-review) similar to those used in schools and universities.\nAverage accu
 racy was 47% and 33% over the base-line while detecting valence and\narous
 al respectively. The result also shows a significant agreement between\npo
 sitive emotions and the user’s engagement.
LOCATION:Rainbow Room (SS03)\, Computer Laboratory
END:VEVENT
END:VCALENDAR