Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | ICPRAM 2014 |
| Untertitel | Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods |
| Seiten | 385-392 |
| Seitenumfang | 8 |
| Publikationsstatus | Veröffentlicht - 2014 |
| Veranstaltung | 3rd International Conference on Pattern Recognition Applications and Methods, ICPRAM 2014 - Angers, Loire Valley, Frankreich Dauer: 6 März 2014 → 8 März 2014 |
Publikationsreihe
| Name | ICPRAM 2014 - Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods |
|---|
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.
ASJC Scopus Sachgebiete
- Informatik (insg.)
- Maschinelles Sehen und Mustererkennung
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
ICPRAM 2014: Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods. 2014. S. 385-392 (ICPRAM 2014 - Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Motion binary patterns for action recognition
AU - Baumann, Florian
AU - Lao, Jie
AU - Ehlers, Arne
AU - Rosenhahn, Bodo
PY - 2014
Y1 - 2014
N2 - 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.
AB - 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.
KW - Human action recognition
KW - IXMAS
KW - KTH
KW - Machine learning
KW - Random forest
KW - Volume local binary patterns
KW - Weizman
UR - http://www.scopus.com/inward/record.url?scp=84902303571&partnerID=8YFLogxK
U2 - 10.5220/0004816903850392
DO - 10.5220/0004816903850392
M3 - Conference contribution
AN - SCOPUS:84902303571
SN - 9789897580185
T3 - ICPRAM 2014 - Proceedings of the 3rd International Conference on Pattern Recognition Applications and Methods
SP - 385
EP - 392
BT - ICPRAM 2014
T2 - 3rd International Conference on Pattern Recognition Applications and Methods, ICPRAM 2014
Y2 - 6 March 2014 through 8 March 2014
ER -