Computer Laboratory

Face Tracking

Tadas Baltrušaitis & Peter Robinson

Examples of tracked faces using our tracker
Examples of tracked faces using our system

Facial expression and head pose are rich sources of information which provide a very important communication channel in our everyday lives. Humans use them to reveal intent, display affection, express emotion, and help regulate turn-taking during conversation. Automated tracking and analysis of such visual cues would greatly benefit human computer interaction.

A crucial initial step in many emotion sensing, face recognition and human behaviour understanding systems is the estimation of head pose and tracking of facial feature points. Tracking these points of interest allows us to analyse their structure and motion. This allows us to understand the underlying meaning of the expressions.

In our work we explore various ways of tracking facial expressions in challenging environments, such as varying head pose, non-uniform lighting, and subtle naturalistic expressions.

The availability of accurate and affordable depth cameras leads us to explore this new and exciting channel for facial expression tracking.

Selected publications

  • Continuous Conditional Neural Fields for Structured Regression
    Tadas Baltrušaitis, Peter Robinson, and Louis-Philippe Morency
    in European Conference on Computer Vision (ECCV), Zürich, Switzerland, September 2014
    [pdf][CCNF code][CLM code]
  • Constrained Local Neural Fields for robust facial landmark detection in the wild
    Tadas Baltrušaitis, Peter Robinson, and Louis-Philippe Morency
    in IEEE Int. Conference on Computer Vision Workshops, 300 Faces in-the-Wild Challenge, Sydney, Australia, December 2013
    [pdf][CLM code]
  • 3D Constrained local model for rigid and non-rigid facial tracking
    Tadas Baltrušaitis, Peter Robinson, and Louis-Philippe Morency
    in IEEE Conference on Computer Vision and Pattern Recognition, Providence, RI, June 2012 [project][pdf]
  • Analysis of Colour Space Transforms for Person Independent AAMs
    Tadas Baltrušaitis and Peter Robinson
    in Proceedings of the ACM / SSPNet 2nd International Symposium on Facial Analysis and Animation