AUTOMATIC KERNEL WIDTH SELECTION FOR NEURAL NETWORK BASED VIDEO OBJECT SEGMENTATION

Dubravko Culibrk, Daniel Socek, Oge Marques, Borko Furht

Abstract

Background modelling Neural Networks (BNNs) represent an approach to motion based object segmentation in video sequences. BNNs are probabilistic classifiers with nonparametric, kernel-based estimation of the underlying probability density functions. The paper presents an enhancement of the methodology, introducing automatic estimation and adaptation of the kernel width. The proposed enhancement eliminates the need to determine kernel width empirically. The selection of a kernel-width appropriate for the features used for segmentation is critical to achieving good segmentation results. The improvement makes the methodology easier to use and more adaptive, and facilitates the evaluation of the approach.

Download


Paper Citation


in Harvard Style

Culibrk D., Socek D., Marques O. and Furht B. (2007). AUTOMATIC KERNEL WIDTH SELECTION FOR NEURAL NETWORK BASED VIDEO OBJECT SEGMENTATION . In Proceedings of the Second International Conference on Computer Vision Theory and Applications - Volume 2: VISAPP, ISBN 978-972-8865-74-0, pages 472-479. DOI: 10.5220/0002058704720479


in Bibtex Style

@conference{visapp07,
author={Dubravko Culibrk and Daniel Socek and Oge Marques and Borko Furht},
title={AUTOMATIC KERNEL WIDTH SELECTION FOR NEURAL NETWORK BASED VIDEO OBJECT SEGMENTATION},
booktitle={Proceedings of the Second International Conference on Computer Vision Theory and Applications - Volume 2: VISAPP,},
year={2007},
pages={472-479},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0002058704720479},
isbn={978-972-8865-74-0},
}


in EndNote Style

TY - CONF
JO - Proceedings of the Second International Conference on Computer Vision Theory and Applications - Volume 2: VISAPP,
TI - AUTOMATIC KERNEL WIDTH SELECTION FOR NEURAL NETWORK BASED VIDEO OBJECT SEGMENTATION
SN - 978-972-8865-74-0
AU - Culibrk D.
AU - Socek D.
AU - Marques O.
AU - Furht B.
PY - 2007
SP - 472
EP - 479
DO - 10.5220/0002058704720479