A Robust Pedestrian Detection Approach Based on Support Vector Machine Using Convex Hull

Abstract:

The support vector machine (SVM), an assuring new method for the classification, has been widely used in many areas efficiently. However, the online learning issue of SVM is still not addressed satisfactorily since when a new sample arrives to retrain the SVM to adjust the classifier. This may not be feasible for real-time applications due to the expensive computation cost for re-training the SVM. This paper propose an Online SVM classifier algorithm known as OSVM-CH, which is based on the convex hull vertices selection depends on geometrical features of SVM. From the theoretical point of view, the first d+1(d is the dimension of the input samples) selected samples are proved to be vertices of the convex hull. This guarantees that the selected samples in our method keep the greatest amount of information of the convex hull. From the pedestrian detection application point of view, the new algorithm can update the classifier without reducing its classification performance.

• Periodical: Applied Mechanics and Materials (Volume 573)
• Pages: 501-507
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.573.501
• Online since: June 2014
• Authors: M.K. Revathi*, P. Annapandi, K.P. Ramya
• Keywords: Kernel, Machine Learning, Online Classifier, Online Learning, Pedestrian Detection, Support Vector Machine (SVM)

References

[1] Anguita. D, Ghio. A, Oneto. L, and Ridella. S, In-sample and out of- sample model selection and error estimation for support vector machines, IEEE Trans. Neural Netw. Learn. Syst., vol. 23, no. 9, p.1390–1406, September, (2012).

DOI: 10.1109/tnnls.2012.2202401

[2] Agarwal. S, Saradhib. V, and Karnick. H, Kernel-based online machine learning and support vector reduction, Neurocomputing, vol. 71, nos. 7–9, p.1230–1237, (2008).

DOI: 10.1016/j.neucom.2007.11.023

[3] Barber.C. B and Dobkin.D. P, The Quickhull Algorithm for Convex Hulls, ACM Transactions on Mathematical Software, Vol. 22, No. 4, Pages 469–483, December, (1996).

DOI: 10.1145/235815.235821

[4] Bottou. L, Online Learning and Stochastic Approximations, Cambridge, U.K.: Cambridge Univ. Press, p.9–42, (1998).

[5] Bordes. A, Ertekin. S, Weston. J, and Bottou. L, Fast kernel classifiers with online and active learning, J. Mach. Learn. Res., vol. 6, p.1579–1619, September, (2005).

[6] Bordes. A and Bottou. L, The Huller: A simple and efficient online SVM, in Proc. 16th Eur. Conf. Mach. Learn., Oct. 2005, p.505–512, October, (2005).

DOI: 10.1007/11564096_48

[7] Cheng. S and Shih. F, An improved incremental training algorithm for support vector machines using active query, Pattern Recognit., vol. 40, no. 3, p.964–971, (2007).

DOI: 10.1016/j.patcog.2006.06.016

[8] Geebelen. D, Suykens J.A. K, and Vandewalle. J, Reducing the number of support vectors of SVM classifiers using the smoothed separable case approximation, IEEE Trans. Neural Netw. Learn. Syst, vol. 23, no. 4, p.682–688, April, (2012).

DOI: 10.1109/tnnls.2012.2186314

[9] Menon. A, Large-Scale Support Vector Machines: Algorithms and Theory, [Online]. Available: http: /cseweb. ucsd. edu/∼akmenon/ResearchExam. pdf, (2009).

[10] Qiao. H, Zhang. P, Wang.D., and Zhang. B, An explicit nonlinear mapping for manifold learning, IEEE Trans. Cybern., vol. 43, no. 1, p.51–63, Feb. (2013).

DOI: 10.1109/tsmcb.2012.2198916