Kernel Local Descriptors with Implicit Rotation Matching

Andrei Bursuc, Giorgos Tolias, Hervé Jégou

Abstract: In this work we design a kernelized local feature descriptor and propose a matching scheme for aligning patches quickly and automatically. We analyze the SIFT descriptor from a kernel view and identify and reproduce some of its underlying benefits. We overcome the quantization artifacts of SIFT by encoding pixel attributes in a continous manner via explicit feature maps. Experiments performed on the patch dataset of Brown et al. [3] show the superiority of our descriptor over methods based on supervised learning.