Multi-Eigenspace Learning for Video-Based Face Recognition.

Conference Paper · January 2007
DOI: 10.1007/978-3-540-74549-5_20 · Source: DBLP
Conference: Advances in Biometrics, International Conference, ICB 2007, Seoul, Korea, August 27-29, 2007, Proceedings

    Abstract

    In this paper, we propose a novel online learning method called Multi-Eigenspace Learning which can learn appearance models
    incrementally from a given video stream. For each subject, we try to learn a few eigenspace models using IPCA (Incremental
    Principal Component Analysis). In the process of Multi-Eigenspace Learning, each eigenspace generally contains more and more
    samples except one eigenspace which contains the least number of samples. Then, these learnt eigenspace models are used for
    video-based face recognition. Experimental results show that the proposed method can achieve high recognition rate.