Abstract

A new class of kernels for object recognition based on local image feature representations are introduced in this paper. Formal proofs are given to show that these kernels satisfy the Mercer condition. In addition, multiple types of local features and semilocal constraints are incorporated. Experimental results of SVM classifiers coupled with the proposed kernels are reported on recognition tasks with the COIL-100 database and compared with existing methods. The proposed kernels achieved competitive performance and were robust to changes in object configurations and image degradations.

Citations

6696	Statistical Learning Theory - Vapnik - 1998
3104	Distinctive Image Features from Scale-Invariant Keypoints - Lowe - 2004
2038	LIBSVM: a library for support vector machines, 2001. Software available at http://www.csie.ntu.edu.tw/˜cjlin/libsvm - Chang, Lin
1252	A combined corner and edge detector - Harris, Stephens - 1988
774	A performance evaluation of local descriptors - Mikolajczyk, Schmid - 2003
519	Kernel methods for pattern analysis - Shawe-Taylor, Cristianini - 2004
367	Local gray-value invariants for image retrieval - Schmid, Mohr - 1997
245	Indexing based on scale invariant interest points - Mikolajczyk, Schmid - 2001
224	Evaluation of interest point detectors - Schmid, Mohr, et al.
143	Support vector machines for 3D object recognition - Pontil, Verri - 1998
118	Columbia Object Image Library (COIL-100 - Nene, Nayar, et al. - 1996
90	Recognition with local features: the kernel recipe - Wallraven, Caputo, et al. - 2003
77	Viewpoint invariant texture matching and wide baseline stereo - Schaffalitzky, Zisserman
68	A kernel between sets of vectors - Kondor, Jebara - 2003
58	A Kullback-Leibler divergence based kernel for SVM classification in multimedia applications - Moreno, Ho, et al. - 2003
26	Evaluation of salient point techniques - Sebe, Tian, et al. - 2004
25	Phase-based local features - Carneiro, Jepson - 2002
23	Building kernels from binary strings for image matching - Odone, Barla, et al.
19	Object categorization with SVM: kernels for local features,” Max Planck Institute for Biological Cybernetics - Eichhorn, Chapelle - 2004
17	SVMs for histogram based image classification - Chapelle, Haffner, et al. - 1999
13	Non-mercer kernel for svm object recognition - Boughorbel, Tarel, et al. - 2004
2	Convolution kernel for structure data - Haussler - 1999
2	Kernel principle angles for classifiication machines with applications to image sequence interpretation - Wolf, Shashua - 2003