Motion Binary Patterns for Action Recognition

Florian Baumann, Jie Lao, Arne Ehlers, Bodo Rosenhahn

2014

Abstract

In this paper, we propose a novel feature type to recognize human actions from video data. By combining the benefit of Volume Local Binary Patterns and Optical Flow, a simple and efficient descriptor is constructed. Motion Binary Patterns (MBP) are computed in spatio-temporal domain while static object appearances as well as motion information are gathered. Histograms are used to learn a Random Forest classifier which is applied to the task of human action recognition. The proposed framework is evaluated on the well-known, publicly available KTH dataset, Weizman dataset and on the IXMAS dataset for multi-view action recognition. The results demonstrate state-of-the-art accuracies in comparison to other methods.

Paper Citation

in Harvard Style

Baumann F., Lao J., Ehlers A. and Rosenhahn B. (2014). Motion Binary Patterns for Action Recognition . In Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-018-5, pages 385-392. DOI: 10.5220/0004816903850392

in Bibtex Style

@conference{icpram14,
author={Florian Baumann and Jie Lao and Arne Ehlers and Bodo Rosenhahn},
title={Motion Binary Patterns for Action Recognition},
booktitle={Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2014},
pages={385-392},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0004816903850392},
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 - Motion Binary Patterns for Action Recognition
SN - 978-989-758-018-5
AU - Baumann F.
AU - Lao J.
AU - Ehlers A.
AU - Rosenhahn B.
PY - 2014
SP - 385
EP - 392
DO - 10.5220/0004816903850392