What energy functions can be minimized via graph cuts?

By Vladimir Kolmogorov, Ramin Zabih. 2002

3,738 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

3D shape segmentation via shape fully convolutional networks (2017, methodology): “use the multi-label graph cut optimisation method [14, 15, 16] to optimise the segmentation results of testing shapes, based on the complementarity between features.”
CloudCV: Large-Scale Distributed Computer Vision as a Cloud Service (2015, background): “Popular examples include object detection [30], articulated body-pose estimation [51], graph-cuts for image segmentation [33], etc.”
An experimental comparison of min-cut/max- flow algorithms for energy minimization in vision (2001, methodology): “This work was mainly done while the authors were with Siemens Corp. Research, Princeton, NJ.”
Learning with Submodular Functions: A Convex Optimization Perspective (2011, neutral): “, [84] and references therein).”
Initial results in underwater single image dehazing (2010, methodology): “3 shows the depth map generation process where (b) shows the initial estimate of the transmission produced by our prior and (c) shows the transmission estimate after being refined using the natural image matting method proposed by Levin et al [12].”
Perturb-and-MAP random fields: Using discrete optimization to learn and sample from energy models (2011, background): “While the order-1 perturbation preserves submodularity [17], order-2 perturbation can yield non-submodular functions even when the original energy is submodular.”
Stacked Hierarchical Labeling (2010, background): “Furthermore, even with simplified representations exact inference remains intractable [13] and prohibits learning these models.”
LOCUS: learning object classes with unsupervised segmentation (2005, methodology): “We can find the deformation field D which maximises (10) by using graph cuts in an α-expansion algorithm, as described in [18, 19].”
NON-ASSOCIATIVE MARKOV NETWORKS FOR 3D POINT CLOUD CLASSIFICATION (2010, methodology): “The problem of minimizing energy of this form is NP-hard whenever K > 2 (Kolmogorov and Zabih, 2004), but there exist a number of methods for finding an approximate minimum.”
Piecewise Flat Embedding for Image Segmentation (2018, methodology): “Graph cuts [35] solves the partition problem using an energy function consisting of a posterior term and a Potts model.”
GeoS: Geodesic Image Segmentation (2008, background): “General energy models of this kind have not been used much in the literature because of the lack of appropriate optimization techniques [25].”
Minimizing Nonsubmodular Functions with Graph Cuts-A Review (2007, background): “There is an accepted view within the computer vision community popularized by [24] that graph cuts can only be used for minimizing submodular energy functions, i.”
Structured Learning and Prediction in Computer Vision (2011, methodology): “Details are given by Kolmogorov and Zabin [80].”
Continuous Global Optimization in Multiview 3D Reconstruction (2007, background): “Corresponding energy functionals (see (3)) describe a typical space partitioning problem and can be also optimized globally in a spatially discrete setting via graph cuts (Kolmogorov and Zabih, 2002).”