BOLD - Binary Online Learned Descriptor For Efficient Image Matching

Vassileios Balntas, Lilian Tang, Krystian Mikolajczyk; The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015, pp. 2367-2375

Abstract


In this paper we propose a novel approach to generate a binary descriptor optimized for each image patch independently. The approach is inspired by the linear discriminant embedding that simultaneously increases inter and decreases intra class distances. A set of discriminative and uncorrelated binary tests is established from all possible tests in an offline training process. The patch adapted descriptors are then efficiently built online from a subset of tests which lead to lower intra class distances thus a more robust descriptor. A patch descriptor consists of two binary strings where one represents the results of the tests and the other indicates the subset of the patch-related robust tests that are used for calculating a masked Hamming distance. Our experiments on three different benchmarks demonstrate improvements in matching performance, and illustrate that per-patch optimization outperforms global optimization.

Related Material


[pdf]
[bibtex]
@InProceedings{Balntas_2015_CVPR,
author = {Balntas, Vassileios and Tang, Lilian and Mikolajczyk, Krystian},
title = {BOLD - Binary Online Learned Descriptor For Efficient Image Matching},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2015}
}