Nonlinear Shape and Appearance Models for Facial Expression Analysis and Synthesis

Authors:
Chan-Su Lee;Ahmed Elgammal
Affiliations:
Rutgers University;Rutgers University
Venue:
ICPR '06 Proceedings of the 18th International Conference on Pattern Recognition - Volume 01
Year:
2006

Citing 0
Cited 3

Facial expression recognition in still pictures and videos using active appearance models: a comparison approach

CompSysTech '07 Proceedings of the 2007 international conference on Computer systems and technologies
Integrating a mixed-feature model and multiclass support vector machine for facial expression recognition

Integrated Computer-Aided Engineering
A realistic dynamic facial expression transfer method

Neurocomputing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Facial expression passes through nonlinear shape and appearance deformations with variations in different people and expressions. We present nonlinear shape and appearance models for facial expression analysis and synthesis using nonlinear generative models for different facial expressions in different people. To achieve accurate shape normalized appearance models, we utilize nonlinear warping using thin plate spline (TPS). A novel nonlinear generative model using conceptual manifold embedding and empirical kernel maps for facial expressions provides facial shape and appearance samples according to the configuration, personal style, and expression parameters. We can recognize facial expressions based on estimated facial expression parameters after iterative estimations of facial expression and style. In addition, the model provides accurate synthesis of facial expression sequences even with high nonlinear deformations of shape and appearance during facial expressions.