We propose a multiple-kernel local-patch descriptor based on efficient match kernels of patch gradients. It combines two parametrizations of gradient position and direction, each parametrization provides robustness to a different type of patch miss-registration: polar parametrization for noise in the patch dominant orientation detection, Cartesian for imprecise location of the feature point. Even though handcrafted, the proposed method consistently outperforms the state-of-the-art methods on two local patch benchmarks.
Cite
@article{arxiv.1707.07825,
title = {Multiple-Kernel Local-Patch Descriptor},
author = {Arun Mukundan and Giorgos Tolias and Ondrej Chum},
journal= {arXiv preprint arXiv:1707.07825},
year = {2017}
}
Comments
To appear in the British Machine Vision Conference (BMVC), September 2017