. Author manuscript; available in PMC: 2017 Feb 1.

Published in final edited form as: IEEE Trans Pattern Anal Mach Intell. 2015 Sep 23;38(8):1640–1650. doi: 10.1109/TPAMI.2015.2481404

TABLE 1.

Classification accuracies on the Auslan database for the proposed approach as well as several state-of-the-art methods. OSCM: Order-preserving Sparse Coding Method. TGAK: Triangular Global Alignment Kernels. RW: Random Walk. RKLR: Regularized Kernel Logistic Regression. Also shown are the computational times of three of the algorithms for which the code is available.

Method	Accuracy	Time/Sample
OSCM [34]	70.46% ± 1.73%	9.43 s
TGAK [10]	88% ± 0.5%	14.35 s
RW Kernel + SVM [24]	89.70% ± 1.78%	2.27 s
Path Kernel + RKLR (l₁)	89.69% ± 1.73%	–
Path Kernel + RKLR (l₂)	90.42% ± 0.67%	–
Path Kernel + SVM	91.60% ± 1.50%	1.62 s