TY - GEN
T1 - A heterogeneous dictionary model for representation and recognition of human actions
AU - Anirudh, Rushil
AU - Ramamurthy, Karthikeyan
AU - Thiagarajan, Jayaraman J.
AU - Turaga, Pavan
AU - Spanias, Andreas
N1 - Funding Information:
This project (IIS-488) was supported by the Investigator-Initiated Study Program of Biosense Webster, Inc., and the Winkelman Family Fund in Cardiovascular Innovation. Dr. Marchlinski has received research support from Biosense Webster, Inc. Dr. Nazarian has received research support from Biosense Webster, ImriCor, and Siemens; and is a consultant for CardioSolv and Circle Software. Dr. Tschabrunn has received research support from Biosense Webster, Attune Medical, and Baylis Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
PY - 2013/10/18
Y1 - 2013/10/18
N2 - In this paper, we consider low-dimensional and sparse representation models for human actions, that are consistent with how actions evolve in high-dimensional feature spaces. We first show that human actions can be well approximated by piecewise linear structures in the feature space. Based on this, we propose a new dictionary model that considers each atom in the dictionary to be an affine subspace defined by a point and a corresponding line. When compared to centered clustering approaches such as K-means, we show that the proposed dictionary is a better generative model for human actions. Furthermore, we demonstrate the utility of this model in efficient representation and recognition of human activities that are not available in the training set.
AB - In this paper, we consider low-dimensional and sparse representation models for human actions, that are consistent with how actions evolve in high-dimensional feature spaces. We first show that human actions can be well approximated by piecewise linear structures in the feature space. Based on this, we propose a new dictionary model that considers each atom in the dictionary to be an affine subspace defined by a point and a corresponding line. When compared to centered clustering approaches such as K-means, we show that the proposed dictionary is a better generative model for human actions. Furthermore, we demonstrate the utility of this model in efficient representation and recognition of human activities that are not available in the training set.
KW - Activity analysis
KW - Dictionary learning
KW - Sparse representations
UR - http://www.scopus.com/inward/record.url?scp=84890492168&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84890492168&partnerID=8YFLogxK
U2 - 10.1109/ICASSP.2013.6638303
DO - 10.1109/ICASSP.2013.6638303
M3 - Conference contribution
AN - SCOPUS:84890492168
SN - 9781479903566
T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
SP - 3472
EP - 3476
BT - 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings
T2 - 2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013
Y2 - 26 May 2013 through 31 May 2013
ER -