Online Learning of Linear Predictors for Real-Time Tracking

Stefan Holzer¹, Marc Pollefeys², Slobodan Ilic¹, David Joseph Tan¹, and Nassir Navab¹

¹Department of Computer Science, Technische Universität München (TUM), Boltzmannstrasse 3, 85748, Garching, Germany
holzers@in.tum.de
slobodan.ilic@in.tum.de
tanda@in.tum.de
navab@in.tum.de

²Department of Computer Science, ETH Zurich, CNB G105 Universitatstrasse 6, CH-8092, Zurich, Switzerland
marc.pollefeys@inf.ethz.ch

Abstract. Although fast and reliable, real-time template tracking using linear predictors requires a long training time. The lack of the ability to learn new templates online prevents their use in applications that require fast learning. This especially holds for applications where the scene is not known a priori and multiple templates have to be added online. So far, linear predictors had to be either learned offline [1] or in an iterative manner by starting with a small sized template and growing it over time [2]. In this paper, we propose a fast and simple reformulation of the learning procedure that allows learning new linear predictors online.

Keywords: template tracking, template learning, linear predictors

LNCS 7572, p. 470 ff.

Full article in PDF | BibTeX

lncs@springer.com
© Springer-Verlag Berlin Heidelberg 2012