Gauss–Laguerre wavelet textural feature fusion with geometrical information for facial expression identification

Poursaberi Ahmad, Noubari Hossein, Gavrilova Marina, Yanushkevich

Description

Facial expressions are a valuable source of information that accompanies facial biometrics. Early detection of physiological and psycho-emotional data from facial expressions is linked to the situational awareness module of any advanced biometric system for personal state re/identification. In this article, a new method that utilizes both texture and geometric information of facial fiducial points is presented. We investigate Gauss–Laguerre wavelets, which have rich frequency extraction capabilities, to extract texture information of various facial expressions. Rotation invariance and the multiscale approach of these wavelets make the feature extraction robust. Moreover, geometric positions of fiducial points provide valuable information for upper/lower face action units. The combination of these two types of features is used for facial expression classification. The performance of this system has been validated on three public databases: the JAFFE, the Cohn-Kanade, and the MMI image.

Sujets

• Facial expression

Informations

• Publié par: biomed
• Publié le: 01 janvier 2012
• Nombre de lectures: 49
• Langue: English
• Poids de l'ouvrage: 2 Mo

Extrait

Poursaberi et al. EURASIP Journal on Image and Video Processing 2012, 2012 :17 http://jivp.eurasipjournals.com/content/2012/1/17

R E S E A R C H Open Access Gauss – Laguerre wavelet textural feature fusion with geometrical information for facial expression identification Ahmad Poursaberi 1* , Hossein Ahmadi Noubari 2 , Marina Gavrilova 1 and Svetlana N Yanushkevich 1

Abstract Facial expressions are a valuable source of information that accompanies facial biometrics. Early detection of physiological and psycho-emotional data from facial expressions is linked to the situational awareness module of any advanced biometric system for personal state re/identification. In this article, a new method that utilizes both texture and geometric information of facial fiducial points is presented. We investigate Gauss – Laguerre wavelets, which have rich frequency extraction capabilities, to extract texture information of various facial expressions. Rotation invariance and the multiscale approach of these wavelets make the feature extraction robust. Moreover, geometric positions of fiducial points provide valuable information for upper/lower face action units. The combination of these two types of features is used for facial expression classification. The performance of this system has been validated on three public databases: the JAFFE, the Cohn-Kanade, and the MMI image. Keywords: Facial expression, Gauss – Laguerre wavelet, Feature fusion, Texture analysis

Introduction taken in 1978 by Suwa et al. [2] who presented a system Automatic facial expression recognition (AFER) is of for facial expression analysis from video, and tracking 20 interest to researchers because of its importance for fa- points as features. Before that, only two ways existed for cial biometric-based intelligent support systems. It pro- FER [3]: (a) human observer-based coding system which vides a behavioral measure to assess emotions, cognitive is subjective, time-consuming, and hard to standardize, processes, and social interaction [1]. Examples of appli- and (b) electromyography-based systems which is inva- cations of AFER include robotics, human – computer sive (needs sensors on the face). The muscle actions re- interface, behavioral science, animations and computer sult in various facial behaviors and motions, and later on games, educational software, emotion processing, and fa- can be used to represent the corresponding facial tigue detection. Due to multiple limitations and difficul- expressions. These assumptions became the basis for ties such as occlusion, lighting conditions, and variation developing the following systems for coding multiple fa- of expressions across the population, or even for an indi- cial expressions and emotions: vidual, having an automatic system helps in creating in- telligent visual media for understanding different 1 The Facial Action Coding System (FACS) — Ekman expressions. Moreover, this understanding helps in and Friesen [4]. building meaningful and responsive HCI interfaces. 2 The Facial Animation parameters (FAPs) — MPEG-4 Each AFER implements three main functions: face de-standard, SNHC [5]. tection and tracking, feature extraction, and expression classification. The first attempt towards the AFER was In the study of Ekman and Friesen [6], it was shown that those six emotions — anger, disgust, fear, happiness, * Correspondence: [e-mail] l s i a t d er n a e t s e s, c a ul n t d ur s e u ” r . p S ri o se me — ti ar m e es “ , di a sc n r e i u m t i r n al ab e l x e pr w e it ss h i i o n n a i n s y co o n n - e 1 Department of Electrical and Computer Engineering, University of Calgary, Canada ENA221, ICT Building, 2500 University Drive NW, Calgary T2N 1N4, sidered as a seventh expression. The FACS describes fa- F C u a ll n l a is d t a ofauthorinformationisavailableattheendofthearticle cial expressions in terms of action units (AUs). It © 2012 Poursaberi et al.; licensee Springer. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.