Foreground-background segmentation using iterated distribution matching

Abstract

This paper addresses the problem of image segmentation with a reference distribution. Recent studies have shown that segmentation with global consistency measures outperforms conventional techniques based on pixel-wise measures. However, such global approaches require a precise distribution to obtain the correct extraction. To overcome this strict assumption, we propose a new approach in which the given reference distribution plays a guiding role in inferring the latent distribution and its consistent region. The inference is based on an assumption that the latent distribution resembles the distribution of the consistent region but is distinct from the distribution of the complement region. We state the problem as the minimization of an energy function consisting of global similarities based on the Bhattacharyya distance and then implement a novel iterated distribution matching process for jointly optimizing distribution and segmentation. We evaluate the proposed algorithm on the GrabCut dataset, and demonstrate the advantages of using our approach with various segmentation problems, including interactive segmentation, background subtraction, and co-segmentation.

Pseudo-bound Optimization for Binary Energies

“…In vision, bound optimizers were recently used for high-order or non-submodular pairwise energies [3,29,11,31]. The recent Auxiliary Cuts [3] work derived bounds for certain class of high-order functions.…”

“…Recently high-order [2,3,12,13,15,26,29,38] and non-submodular pairwise [11,16,21,19] energy minimization have drawn tremendous research interests. Those energy functions arise naturally in many computer vision and image processing applications.…”

“…Those energy functions arise naturally in many computer vision and image processing applications. Examples of high-order functions include but are not limited to constraints on segment volume [12,38], clique labeling consistency [13,18,35] and matching target distributions [2,3,12,29]. Pairwise non-submodular energies occur in deconvolution [11], curvature regularization [8,11,28], inpainting [16] and surface registration [16].…”

“…The recent studies [11,12] have shown that FTR achieved the state of the art performance in many applications. Our work is more closely related to bound optimizers, which take auxiliary functions [2,3,11,22,27,29,31] as local approximations and were recently shown to yield competitive performances in several vision problems [3,11,29].…”

“…By relaxing the integrality constraints, they globally transform the original function into a linear function with extra variables and linear constraints. Unlike global linearization, local techniques iteratively approximate the original energy around current solution, for instance, using Taylor approximations [11,12,17,24] or auxiliary functions [2,3,11,27,29,31]. The recent Fast Trust Region (FTR) method [12] finds the optimal solution of a local approximation within a trust region, i.e.…”