Please use this identifier to cite or link to this item: http://hdl.handle.net/10553/43989
Title: Using a discrete Hidden Markov Model Kernel for lip-based biometric identification
Authors: Travieso, Carlos M. 
Zhang, Jianguo
Miller, Paul
Alonso, Jesús B. 
UNESCO Clasification: 3307 Tecnología electrónica
Keywords: Discrete Hidden Markov Model Kernel, Image processing, Lip-based biometrics, Pattern recognition
Issue Date: 2014
Publisher: 0262-8856
Journal: Image and Vision Computing 
Abstract: In this paper, a novel and effective lip-based biometric identification approach with the Discrete Hidden Markov Model Kernel (DHMMK) is developed. Lips are described by shape features (both geometrical and sequential) on two different grid layouts: rectangular and polar. These features are then specifically modeled by a DHMMK, and learnt by a support vector machine classifier. Our experiments are carried out in a ten-fold cross validation fashion on three different datasets, GPDS-ULPGC Face Dataset, PIE Face Dataset and RaFD Face Dataset. Results show that our approach has achieved an average classification accuracy of 99.8%, 97.13%, and 98.10%, using only two training images per class, on these three datasets, respectively. Our comparative studies further show that the DHMMK achieved a 53% improvement against the baseline HMM approach. The comparative ROC curves also confirm the efficacy of the proposed lip contour based biometrics learned by DHMMK. We also show that the performance of linear and RBF SVM is comparable under the frame work of DHMMK.
URI: http://hdl.handle.net/10553/43989
ISSN: 0262-8856
DOI: 10.1016/j.imavis.2014.10.001
Source: Image and Vision Computing[ISSN 0262-8856],v. 32, p. 1080-1089
Appears in Collections:Artículos
Show full item record

SCOPUSTM   
Citations

7
checked on Apr 11, 2021

WEB OF SCIENCETM
Citations

7
checked on Apr 11, 2021

Page view(s)

17
checked on Apr 11, 2021

Google ScholarTM

Check

Altmetric


Share



Export metadata



Items in accedaCRIS are protected by copyright, with all rights reserved, unless otherwise indicated.