Temporal validation of particle filters for video tracking
Entity
UAM. Departamento de Tecnología Electrónica y de las Comunicaciones; QMUL. Centre for Intelligent Sensing (CIS, UK)Publisher
Elsevier B.V.Date
2015-02Citation
10.1016/j.cviu.2014.06.016
Journal Computer Vision and Image Understanding 131 (2015): 42-55
ISSN
1077-3142 (print); 1090-235X (online)DOI
10.1016/j.cviu.2014.06.016Funded by
This work was partially supported by the Spanish Government (EventVideo, TEC2011-25995) and by the EU Crowded Environments monitoring for Activity Understanding and Recognition (CENTAUR, FP7-PEOPLE-2012-IAPP) project under GA number 324359. Most of the work reported in this paper was done at the Centre for Intelligent Sensing in Queen Mary University of London.Project
info:eu-repo/grantAgreement/EC/FP7/324359Editor's Version
http://dx.doi.org/10.1016/j.cviu.2014.06.016Subjects
Change detection; Model validation; Particle Filter; Performance evaluation; Uncertainty; Video tracking; TelecomunicacionesNote
This is the author’s version of a work that was accepted for publication in Journal Computer Vision and Image Understanding. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Journal Computer Vision and Image Understanding, 131 (2015) DOI: 10.1016/j.cviu.2014.06.016Rights
© 2015 Elsevier B.V. All rights reservedEsta obra está bajo una licencia de Creative Commons Reconocimiento-NoComercial-SinObraDerivada 4.0 Internacional.
Abstract
A novel approach to determine adaptively the temporal consistency of Particle Filters.The proposed method is demonstrated on online performance evaluation of tracking.Temporal consistency is modeled by convolutions of mixtures of Gamma distributions.The proposed method does not need thresholds and can be used on large datasets. We present an approach for determining the temporal consistency of Particle Filters in video tracking based on model validation of their uncertainty over sliding windows. The filter uncertainty is related to the consistency of the dispersion of the filter hypotheses in the state space. We learn an uncertainty model via a mixture of Gamma distributions whose optimum number is selected by modified information-based criteria. The time-accumulated model is estimated as the sequential convolution of the uncertainty model. Model validation is performed by verifying whether the output of the filter belongs to the convolution model through its approximated cumulative density function. Experimental results and comparisons show that the proposed approach improves both precision and recall of competitive approaches such as Gaussian-based online model extraction, bank of Kalman filters and empirical thresholding. We combine the proposed approach with a state-of-the-art online performance estimator for video tracking and show that it improves accuracy compared to the same estimator with manually tuned thresholds while reducing the overall computational cost.
Files in this item
Google Scholar:San Miguel Avedillo, Juan Carlos
-
Cavallaro, Andrea
This item appears in the following Collection(s)
Related items
Showing items related by title, author, creator and subject.