Title :
Face recognition methods for multimodal interface
Author :
Ban, Jozef ; Pavlovicova, Jarmila ; Feder, Meir ; Omelina, Lubes ; Oravec, Milos
Author_Institution :
Inst. of Telecommun., Slovak Univ. of Technol., Bratislava, Slovakia
Abstract :
In this paper we provide a comparative study of several conventional face recognition methods (PCA, KPCA, GDA, SVM and RBF) that are suitable to work properly in multimodal systems. Performance of these systems is often influenced by various negative effects of the real-world environment. We evaluate the influence of varying illuminations and pose of faces on face recognition accuracy. Based on the results of our experiments we select the most suitable face recognition methods for application in HBB-NEXT project. This project wants to lay the foundations for advanced hybrid multiuser services.
Keywords :
face recognition; human computer interaction; principal component analysis; radial basis function networks; support vector machines; GDA; HBB-NEXT project; KPCA; PCA; RBF; SVM; face recognition methods; generalized discriminant analysis; kernel principal component analysis; multimodal interface; multimodal systems; principal component analysis; radial basis function neural network; services; support vector machine; Databases; Face; Face recognition; Kernel; Lighting; Principal component analysis; Support vector machines; face recognition; machine learning methods; multimodal interface;
Conference_Titel :
Wireless and Mobile Networking Conference (WMNC), 2012 5th Joint IFIP
Conference_Location :
Bratislava
Print_ISBN :
978-1-4673-2993-4
DOI :
10.1109/WMNC.2012.6416164
