Initialization Framework for Latent Variable Models
Heydar Maboudi Afkham, Carl Henrik Ek, Stefan Carlsson
2014
Abstract
In this paper, we discuss the properties of a class of latent variable models that assumes each labeled sample is associated with set of different features, with no prior knowledge of which feature is the most relevant feature to be used. Deformable-Part Models (DPM) can be seen as good example of such models. While Latent SVM framework (LSVM) has proven to be an efficient tool for solving these models, we will argue that the solution found by this tool is very sensitive to the initialization. To decrease this dependency, we propose a novel clustering procedure, for these problems, to find cluster centers that are shared by several sample sets while ignoring the rest of the cluster centers. As we will show, these cluster centers will provide a robust initialization for the LSVM framework.
DownloadPaper Citation
in Harvard Style
Maboudi Afkham H., Ek C. and Carlsson S. (2014). Initialization Framework for Latent Variable Models . In Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-018-5, pages 227-232. DOI: 10.5220/0004826302270232
in Bibtex Style
@conference{icpram14,
author={Heydar Maboudi Afkham and Carl Henrik Ek and Stefan Carlsson},
title={Initialization Framework for Latent Variable Models},
booktitle={Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2014},
pages={227-232},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0004826302270232},
isbn={978-989-758-018-5},
}
in EndNote Style
TY - CONF
JO - Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - Initialization Framework for Latent Variable Models
SN - 978-989-758-018-5
AU - Maboudi Afkham H.
AU - Ek C.
AU - Carlsson S.
PY - 2014
SP - 227
EP - 232
DO - 10.5220/0004826302270232