Beyond Prototypic Expressions: Discriminating Subtle Changes in the Face - Robotics Institute Carnegie Mellon University

Beyond Prototypic Expressions: Discriminating Subtle Changes in the Face

Jeffrey Cohn, Jenn-Jier James Lien, Takeo Kanade, Wei Hua, and Adena Zlochower
Workshop Paper, 7th IEEE Robot and Human Communications Workshop, September, 1998

Abstract

Current approaches to automated analysis have focused on a small set of prototypic expressions (e.g., joy or anger). Prototypic expressions occur infrequently in everyday life, however, and emotion expression is far more varied. To capture the full range of emotion expression, automated discrimination of fine-grained changes in facial expression is needed. We developed and implemented a computer vision system, Automated Face Analysis, that is sensitive to subtle changes in the face. Three convergent modules extract feature information and discriminate FACS action units using Hidden Markov Models. The modules include feature-point and dense-flow tracking and high-gradient component detection. In image sequences from 100 young adults, action units and action unit combinations in the brow, eye, and mouth regions were selected for analysis if they occurred a minimum of 25 times in the image database. Selected facial features were automatically tracked using hierarchical algorithms to estimate optical flow and high-gradient components. Image sequences were randomly divided into training and test sets. Automated Face Analysis demonstrated high concurrent validity with manual FACS coding.

BibTeX

@workshop{Cohn-1998-16558,
author = {Jeffrey Cohn and Jenn-Jier James Lien and Takeo Kanade and Wei Hua and Adena Zlochower},
title = {Beyond Prototypic Expressions: Discriminating Subtle Changes in the Face},
booktitle = {Proceedings of 7th IEEE Robot and Human Communications Workshop},
year = {1998},
month = {September},
}