TUTCRIS - Tampereen teknillinen yliopisto

TUTCRIS

Distance-based human action recognition using optimized class representations

Tutkimustuotosvertaisarvioitu

Standard

Distance-based human action recognition using optimized class representations. / Iosifidis, Alexandros; Tefas, Anastasios; Pitas, Ioannis.

julkaisussa: Neurocomputing, Vuosikerta 161, 05.08.2015, s. 47-55.

Tutkimustuotosvertaisarvioitu

Harvard

Iosifidis, A, Tefas, A & Pitas, I 2015, 'Distance-based human action recognition using optimized class representations', Neurocomputing, Vuosikerta. 161, Sivut 47-55. https://doi.org/10.1016/j.neucom.2014.10.088

APA

Vancouver

Author

Iosifidis, Alexandros ; Tefas, Anastasios ; Pitas, Ioannis. / Distance-based human action recognition using optimized class representations. Julkaisussa: Neurocomputing. 2015 ; Vuosikerta 161. Sivut 47-55.

Bibtex - Lataa

@article{e7965997f9be4eb9842e5b869bc07c50,
title = "Distance-based human action recognition using optimized class representations",
abstract = "We study distance-based classification of human actions and introduce a new metric learning approach based on logistic discrimination for the determination of a low-dimensional feature space of increased discrimination power. We argue that for effective distance-based classification, both the optimal projection space and the optimal class representation should be determined. We qualitatively and quantitatively illustrate the superiority of the proposed approach to metric learning approaches employing the class mean for class representation. We also introduce extensions of the proposed metric learning approach to allow for richer class representations and to operate in arbitrary-dimensional Hilbert spaces for non-linear feature extraction and classification. Experimental results denote that the performance of the proposed distance-based classification schemes is comparable (or even better) to that of Support Vector Machine classifier (in both the linear and kernel cases) which is currently the standard choice for human action recognition.",
keywords = "Distance-based classification, Optimized class representations",
author = "Alexandros Iosifidis and Anastasios Tefas and Ioannis Pitas",
year = "2015",
month = "8",
day = "5",
doi = "10.1016/j.neucom.2014.10.088",
language = "English",
volume = "161",
pages = "47--55",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier Science B.V.",

}

RIS (suitable for import to EndNote) - Lataa

TY - JOUR

T1 - Distance-based human action recognition using optimized class representations

AU - Iosifidis, Alexandros

AU - Tefas, Anastasios

AU - Pitas, Ioannis

PY - 2015/8/5

Y1 - 2015/8/5

N2 - We study distance-based classification of human actions and introduce a new metric learning approach based on logistic discrimination for the determination of a low-dimensional feature space of increased discrimination power. We argue that for effective distance-based classification, both the optimal projection space and the optimal class representation should be determined. We qualitatively and quantitatively illustrate the superiority of the proposed approach to metric learning approaches employing the class mean for class representation. We also introduce extensions of the proposed metric learning approach to allow for richer class representations and to operate in arbitrary-dimensional Hilbert spaces for non-linear feature extraction and classification. Experimental results denote that the performance of the proposed distance-based classification schemes is comparable (or even better) to that of Support Vector Machine classifier (in both the linear and kernel cases) which is currently the standard choice for human action recognition.

AB - We study distance-based classification of human actions and introduce a new metric learning approach based on logistic discrimination for the determination of a low-dimensional feature space of increased discrimination power. We argue that for effective distance-based classification, both the optimal projection space and the optimal class representation should be determined. We qualitatively and quantitatively illustrate the superiority of the proposed approach to metric learning approaches employing the class mean for class representation. We also introduce extensions of the proposed metric learning approach to allow for richer class representations and to operate in arbitrary-dimensional Hilbert spaces for non-linear feature extraction and classification. Experimental results denote that the performance of the proposed distance-based classification schemes is comparable (or even better) to that of Support Vector Machine classifier (in both the linear and kernel cases) which is currently the standard choice for human action recognition.

KW - Distance-based classification

KW - Optimized class representations

UR - http://www.scopus.com/inward/record.url?scp=84929045315&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2014.10.088

DO - 10.1016/j.neucom.2014.10.088

M3 - Article

VL - 161

SP - 47

EP - 55

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

ER -