National Transport Library Research Database

UN-EYE Avidentifiering av Eyetracker video (Un-identification of Eyetracker video)

Sponsors, duration, budget: Vinnova ; 2009-03-01 -- 2011-01-30 ; 2045890 kronorRegistration number:
  • Vinnova 200900088
Subject(s): Online resources: Abstract: Projektet syftar till att utveckla program, som uttrycker eytracker video i 6-8 generiska FACS-koder, som beskriver ansiktuttrycken (t.ex. glad, trött, upprörd etc.).Abstract: The idea is to use Active Appearance Models (AAM) to track deformations in face space and a neural classifier to map onto the Facial Action Code System (FACS) codes. The FACS represents the facial expressions without the possibility to identify the test subject and form a basis for reanimation of the facial expressions. The project aims to develop the program, expressing eytracker video of 6-8 generic FACS codes, which describe different expression (eg, happy, tired, upset, etc.).
Item type: