DocumentCode :
3510495
Title :
Heteroscedastic probabilistic linear discriminant analysis for manifold learning in video-based face recognition
Author :
Wibowo, Moh Edi ; Tjondronegoro, Dian ; Ligang Zhang ; Himawan, Ivan
Author_Institution :
Sci. & Eng. Fac., Queensland Univ. of Technol., Brisbane, QLD, Australia
fYear :
2013
fDate :
15-17 Jan. 2013
Firstpage :
46
Lastpage :
52
Abstract :
To recognize faces in video, face appearances have been widely modeled as piece-wise local linear models which linearly approximate the smooth yet non-linear low dimensional face appearance manifolds. The choice of representations of the local models is crucial. Most of the existing methods learn each local model individually meaning that they only anticipate variations within each class. In this work, we propose to represent local models as Gaussian distributions which are learned simultaneously using the heteroscedastic probabilistic linear discriminant analysis (PLDA). Each gallery video is therefore represented as a collection of such distributions. With the PLDA, not only the within-class variations are estimated during the training, the separability between classes is also maximized leading to an improved discrimination. The heteroscedastic PLDA itself is adapted from the standard PLDA to approximate face appearance manifolds more accurately. Instead of assuming a single global within-class covariance, the heteroscedastic PLDA learns different within-class covariances specific to each local model. In the recognition phase, a probe video is matched against gallery samples through the fusion of point-to-model distances. Experiments on the Honda and MoBo datasets have shown the merit of the proposed method which achieves better performance than the state-of-the-art technique.
Keywords :
Gaussian distribution; approximation theory; face recognition; image matching; image representation; learning (artificial intelligence); statistical analysis; video signal processing; Gaussian distribution; Honda dataset; MoBo dataset; PLDA; approximation; face appearance; face appearance manifold; face representation; heteroscedastic probabilistic linear discriminant analysis; manifold learning; piecewise local linear model; probe video matching; video-based face recognition; within-class variation; Computational modeling; Face; Hidden Markov models; Manifolds; Probes; Standards; Training;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Applications of Computer Vision (WACV), 2013 IEEE Workshop on
Conference_Location :
Tampa, FL
ISSN :
1550-5790
Print_ISBN :
978-1-4673-5053-2
Electronic_ISBN :
1550-5790
Type :
conf
DOI :
10.1109/WACV.2013.6474998
Filename :
6474998
Link To Document :
بازگشت