Student Topics

Automatic FACS Encoding from Videos

Status of Master Thesis: open
Supervisors: Sebastian Zambanini and Martin Kampel

Problem Statement

The Facial Action Coding System (FACS) is a system to taxonomize human facial movements by their appearance on the face. It is a common standard to systematically categorize the physical expression of emotions which is achieved by assigning specific Action Units (AUs) to temporal segments of a video that show the expression.  It is commonly used by psychologists and animators, but has the drawback of being a time-consuming task when done manually. Therefore, an automatic method would help to save time, but also to achieve a more objective encoding and to reduce the problem of inter-coder variance.

facs

Goal

Based on a literature research, a system for automatic FACS encoding should be developed and evaluated on provided real-world video data.

Workflow

  • Literature research on existing methods for automatic FACS encoding from images/videos
  • Development and evaluation of the system
  • Written thesis (in English) and final presentation

Literature:

Fernando de la Torre, Tomas Simon Kruez, Zara Ambadar and Jeffrey F. Cohn, “FAST-FACS: A Computer-Assisted System to Increase Speed and Reliability of Manual FACS Coding“, 4th International Conference on Affective Computing and Intelligent Interaction, pp. 57-66, 2011.

Marian Stewart Bartlett, Gwen Littlewort, Mark Frank, Claudia Lainscsek, Ian Fasel and Javier Movellan, “Fully Automatic Facial Action Recognition in Spontaneous Behavior“, 7th International Conference on Automatic Face and Gesture Recognition, pp. 223-230, 2006.