Face recognition using wireframe model across facial expressions

Zahid Riaz, Christoph Mayer, Michael Beetz, Bernd Radig

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

Abstract

This paper describes face recognition across facial expressions variations. We focus on an automatic feature extraction technique which is not only efficient but also accurate for person identification. A 3D wireframe model is fitted to face images using a robust objective function. Furthermore, we extract structural and textural information which is coupled with temoral information from the motion of local facial features. The extracted information is combined to form a feature vector descriptor for each image. This set of features has been tested on two databases for face recognition across facial expressions. We use Bayesian Network (BN) and Binary Decision Trees (BDT) as classifiers. The developed system is automatic, real-time capable and efficient.

OriginalspracheEnglisch
TitelBiometric ID Management and Multimodal Communication - Joint COST 2101 and 2102 International Conference, BioID_MultiComm 2009, Proceedings
Seiten122-129
Seitenumfang8
DOIs
PublikationsstatusVeröffentlicht - 2009
VeranstaltungJoint COST 2101 and 2102 International Conference on Biometric ID Management and Multimodal Communication, BioID_MultiComm 2009 - Madrid, Spanien
Dauer: 16 Sept. 200918 Sept. 2009

Publikationsreihe

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Band5707 LNCS
ISSN (Print)0302-9743
ISSN (elektronisch)1611-3349

Konferenz

KonferenzJoint COST 2101 and 2102 International Conference on Biometric ID Management and Multimodal Communication, BioID_MultiComm 2009
Land/GebietSpanien
OrtMadrid
Zeitraum16/09/0918/09/09

Fingerprint

Untersuchen Sie die Forschungsthemen von „Face recognition using wireframe model across facial expressions“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren