Pedestrian Re-identification - Metric Learning using Symmetric Ensembles of Categories

Sateesh Pedagadi, James Orwell, Boghos Boghossian

2015

Abstract

This paper presents a method for pedestrian re-identification, with two novel contributions. Firstly, each element in the target population is classified into one of n categories, using the expected accuracy of the re-identification estimate for this element. A metric for each category is separately trained using a standard (Local Fisher) method. To process a test set, each element is classified into one of the categories, and the corresponding metric is selected and used. The second contribution is the proposal to use a symmetrised distance measure. A standard procedure is to learn a metric using one set as the probe and the other set as the gallery. This paper generalises that procedure by reversing the labels to learn a different metric, and uses a linear (symmetrised) combination of the two. This can be applied in cases for which there are two distinct sets of observations, i.e. from two cameras, e.g. VIPER. Using this publicly available dataset, it is demonstrated how these contributions result in improved re-identification performance.

Paper Citation

in Harvard Style

Pedagadi S., Orwell J. and Boghossian B. (2015). Pedestrian Re-identification - Metric Learning using Symmetric Ensembles of Categories . In Proceedings of the 10th International Conference on Computer Vision Theory and Applications - Volume 2: VISAPP, (VISIGRAPP 2015) ISBN 978-989-758-090-1, pages 274-280. DOI: 10.5220/0005270602740280

in Bibtex Style

@conference{visapp15,
author={Sateesh Pedagadi and James Orwell and Boghos Boghossian},
title={Pedestrian Re-identification - Metric Learning using Symmetric Ensembles of Categories},
booktitle={Proceedings of the 10th International Conference on Computer Vision Theory and Applications - Volume 2: VISAPP, (VISIGRAPP 2015)},
year={2015},
pages={274-280},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0005270602740280},
isbn={978-989-758-090-1},
}

in EndNote Style

TY - CONF
JO - Proceedings of the 10th International Conference on Computer Vision Theory and Applications - Volume 2: VISAPP, (VISIGRAPP 2015)
TI - Pedestrian Re-identification - Metric Learning using Symmetric Ensembles of Categories
SN - 978-989-758-090-1
AU - Pedagadi S.
AU - Orwell J.
AU - Boghossian B.
PY - 2015
SP - 274
EP - 280
DO - 10.5220/0005270602740280