AFECT


AFECT-1, The Automatic Facial Expression Coding Tool (Version 1) is the core expression recognition engine developed at MPT. It automates the Facial Action Coding System (FACS), a comprehensive and scientifically backed approach for numerically representing all possible facial expressions in the human face. It analyzes images frame by frame and provides the following information

  1. Frame by frame information about the observed intensity of up to 30 Facial Action Units from the facial action coding system.
  2. Frame by frame information about he observed intensity of 6 universal expressions of emotion: Anger, Disgust, Fear, Happiness, Sadness, Surprise.
  3. Frame by frame information about the location of the face, eyes, nose and mouth.
  4. Frame by frame information about the 3D orientation of the face: pan, tilt, and roll.
  5. Frame by frame information about accessories: Presence/absence of glasses.

AFECT-1 can run in  real time mode at approximately 10 frame per second on an 2.6 Ghz Intel Core Duo machine. It can also be run in batch mode to analyze large video datasets.