Abstract: In this paper a variety of computer vision algorithms in a context of their application to the problem of study of facial expressions were analyzed. Performance characteristics of computer vision algorithms for tracking markers which were attached to the face in ideal and real conditions were obtained. A list of algorithms’ faults that occurred during experiments was given, and ways to reduce an impact of various factors that influenced the experiments were suggested. The results of experiments were used to model facial expressions on a face of a virtual model of a human head.

Keywords: facial expression, computer vision, reference points.

Link:

FACIAL EXPRESSIONS ANALYSIS BASED ON A COMPUTER VISION ALGORITHMS

Iurii Krak, Anton Ternov, Vladislav Kuznetsov

http://www.foibg.com/ijicp/vol01/ijicp01-04-p03.pdf