Abstract

Correlation fifilters take advantage of specifific properties in the Fourier domain allowing them to be estimated effificiently: O(ND log D) in the frequency domain, versus O(D3 + ND2) spatially where D is signal length, and N is the number of signals. Recent extensions to correlation fifilters, such as MOSSE, have reignited interest of their use in the vision community due to their robustness and attractive computational properties. In this paper we demonstrate, however, that this computational effificiency comes at a cost. Specififically, we demonstrate that only 1D proportion of shifted examples are unaffected by boundary effects which has a dramatic effect on detection/tracking performance. In this paper, we propose a novel approach to correlation fifilter estimation that: (i) takes advantage of inherent computational redundancies in the frequency domain, (ii) dramatically reduces boundary effects, and (iii) is able to implicitly exploit all possible patches densely extracted from training examples during learning process. Impressive object tracking and detection results are presented in terms of both accuracy and computational effificiency