Results 1 
7 of
7
Applications of parametric maxflow in computer vision
"... The maximum flow algorithm for minimizing energy functions of binary variables has become a standard tool in computer vision. In many cases, unary costs of the energy depend linearly on parameter λ. In this paper we study vision applications for which it is important to solve the maxflow problem for ..."
Abstract

Cited by 40 (7 self)
 Add to MetaCart
The maximum flow algorithm for minimizing energy functions of binary variables has become a standard tool in computer vision. In many cases, unary costs of the energy depend linearly on parameter λ. In this paper we study vision applications for which it is important to solve the maxflow problem for different λ’s. An example is a weighting between data and regularization terms in image segmentation or stereo: it is desirable to vary it both during training (to learn λ from ground truth data) and testing (to select best λ using highknowledge constraints, e.g. user input). We review algorithmic aspects of this parametric maximum flow problem previously unknown in vision, such as the ability to compute all breakpoints of λ and corresponding optimal configurations in finite time. These results allow, in particular, to minimize the ratio of some geometric functionals, such as flux of a vector field over length (or area). Previously, such functionals were tackled with shortest path techniques applicable only in 2D. We give theoretical improvements for “PDE cuts ” [5]. We present experimental results for image segmentation, 3D reconstruction, and the cosegmentation problem. 1.
Network Flow Algorithms for Structured Sparsity
"... We consider a class of learning problems that involve a structured sparsityinducing norm defined as the su mof ℓ∞norms over groups of variables. Whereas a lot of effort has been put in developing fast optimization methods when the groups are disjoint or embedded in a specific hierarchical structur ..."
Abstract

Cited by 35 (11 self)
 Add to MetaCart
We consider a class of learning problems that involve a structured sparsityinducing norm defined as the su mof ℓ∞norms over groups of variables. Whereas a lot of effort has been put in developing fast optimization methods when the groups are disjoint or embedded in a specific hierarchical structure, we address here the case of general overlapping groups. To this end, we show that the corresponding optimization problem is related to network flow optimization. More precisely, the proximal problem associated with the norm we consider is dual to a quadratic mincost flow problem. We propose an efficient procedure which computes its solution exactly in polynomial time. Our algorithm scales up to millions of variables, and opens up a whole new range of applications for structured sparse models. We present several experiments on image and video data, demonstrating the applicability and scalability of our approach for various problems.
Convex and network flow optimization for structured sparsity
 JMLR
"... We consider a class of learning problems regularized by a structured sparsityinducing norm defined as the sum of ℓ2 or ℓ∞norms over groups of variables. Whereas much effort has been put in developing fast optimization techniques when the groups are disjoint or embedded in a hierarchy, we address ..."
Abstract

Cited by 15 (5 self)
 Add to MetaCart
We consider a class of learning problems regularized by a structured sparsityinducing norm defined as the sum of ℓ2 or ℓ∞norms over groups of variables. Whereas much effort has been put in developing fast optimization techniques when the groups are disjoint or embedded in a hierarchy, we address here the case of general overlapping groups. To this end, we present two different strategies: On the one hand, we show that the proximal operator associated with a sum of ℓ∞norms can be computed exactly in polynomial time by solving a quadratic mincost flow problem, allowing the use of accelerated proximal gradient methods. On the other hand, we use proximal splitting techniques, and address an equivalent formulation with nonoverlapping groups, but in higher dimension and with additional constraints. We propose efficient and scalable algorithms exploiting these two strategies, which are significantly faster than alternative approaches. We illustrate these methods with several problems such as CUR matrix factorization, multitask learning of treestructured dictionaries, background subtraction in video sequences, image denoising with wavelets, and topographic dictionary learning of natural image patches.
Experimental evaluation of parametric maxflow algorithms
 In WEA ’07: Proceedings of the 6th Workshop on Experimental Algorithms
, 2007
"... Abstract. The parametric maximum flow problem is an extension of the classical maximum flow problem in which the capacities of certain arcs are not fixed but are functions of a single parameter. Gallo et al. [6] showed that certain versions of the pushrelabel algorithm for ordinary maximum flow can ..."
Abstract

Cited by 6 (1 self)
 Add to MetaCart
Abstract. The parametric maximum flow problem is an extension of the classical maximum flow problem in which the capacities of certain arcs are not fixed but are functions of a single parameter. Gallo et al. [6] showed that certain versions of the pushrelabel algorithm for ordinary maximum flow can be extended to the parametric problem while only increasing the worstcase time bound by a constant factor. Recently Zhang et al. [14,13] proposed a novel, simple balancing algorithm for the parametric problem on bipartite networks. They claimed good performance for their algorithm on networks arising from a realworld application. We describe the results of an experimental study comparing the performance of the balancing algorithm, the GGT algorithm, and a simplified version of the GGT algorithm, on networks related to those of the application of Zhang et al. as well as networks designed to be hard for the balancing algorithm. Our implementation of the balancing algorithm beats both versions of the GGT algorithm on networks related to the application, thus supporting the observations of Zhang et al. On the other hand, the GGT algorithm is more robust; it beats the balancing algorithm on some natural networks, and by asymptotically increasing amount on networks designed to be hard for the balancing algorithm. 1
The Partial Augment–Relabel Algorithm for the Maximum Flow Problem
 In Proc. 16th Annual European Symposium Algorithms
"... Abstract. The maximum flow problem is a classical optimization problem with many applications. For a long time, HIPR, an efficient implementation of the highestlabel pushrelabel algorithm, has been a benchmark due to its robust performance. We propose another variant of the pushrelabel method, t ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
Abstract. The maximum flow problem is a classical optimization problem with many applications. For a long time, HIPR, an efficient implementation of the highestlabel pushrelabel algorithm, has been a benchmark due to its robust performance. We propose another variant of the pushrelabel method, the partial augmentrelabel (PAR) algorithm. Our experiments show that PAR is very robust. It outperforms HIPR on all problem families tested, asymptotically in some cases. 1