Research Institute of Visual Computing

Select a project of interest using the advanced search:

A comparison of the reproducibility of verbal and nonverbal facial gestures using three-dimensional motion analysis

Authors: Popat H, Henley E, Richmond S, Benedikt L, Marshall D, Rosin PL.

DOI: 142(6):867-72.

SUBJECTS AND METHODS:

Twenty-five white subjects were asked to perform four verbal gestures and two nonverbal facial gestures in a normal, relaxed manner. The sequences were captured using a noninvasive, three-dimensional motion analysis scanner (3dMDFace Dynamic System; 3Q Technologies, Atlanta, GA) at an initial time point (T1) and one month later (T2). Principal component analysis was used to analyze the dynamics of lip movement, and dynamic time warping was used to evaluate the reproducibility between T1 and T2 for the individual facial gestures. Statistical analyses were used to compare the reproducibility of the different facial gestures.

RESULTS:

All four verbal gestures were found to be significantly more reproducible than the nonverbal gestures. The word "puppy" was most reproducible, and the normal smile expression least reproducible. A hierarchy of reproducibility is proposed as "puppy," "baby," "rope," "bob," standardized smile, normal smile.

CONCLUSION:

Verbal facial gestures, in particular the words "puppy" and "baby," are more appropriate for use in the assessment of lip movement when compared with smile expressions because of their high level of reproducibility over a one-month period.

Copyright 2010 American Academy of Otolaryngology-Head and Neck Surgery Foundation. Published by Mosby, Inc. All rights reserved.

Authors

Prof. David Marshall

Prof. David Marshall

Computer Vision including Reverse Engineering, Automated Inspection, Dynamic 2D/3D Facial Analysis.

Prof. Paul Rosin

Prof. Paul Rosin

Various aspects of computer vision, including 2D and 3D facial analysis and synthesis.