Monotone Boolean Functions, Feasibility/Infeasibility, LP-type problems and MaxCon (original) (raw)
Related papers
Consensus Maximisation Using Influences of Monotone Boolean Functions
2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Consensus maximisation (MaxCon), which is widely used for robust fitting in computer vision, aims to find the largest subset of data that fits the model within some tolerance level. In this paper, we outline the connection between MaxCon problem and the abstract problem of finding the maximum upper zero of a Monotone Boolean Function (MBF) defined over the Boolean Cube. Then, we link the concept of influences (in a MBF) to the concept of outlier (in MaxCon) and show that influences of points belonging to the largest structure in data would generally be smaller under certain conditions. Based on this observation, we present an iterative algorithm to perform consensus maximisation. Results for both synthetic and real visual data experiments show that the MBF based algorithm is capable of generating a near optimal solution relatively quickly. This is particularly important where there are large number of outliers (gross or pseudo) in the observed data.
Maximum Consensus by Weighted Influences of Monotone Boolean Functions
arXiv (Cornell University), 2021
Robust model fitting is a fundamental problem in computer vision: used to pre-process raw data in the presence of outliers. Maximisation of Consensus (MaxCon) is one of the most popular robust criteria and widely used. Recently (Tennakoon et al. CVPR2021), a connection has been made between MaxCon and estimation of influences of a Monotone Boolean function. Equipping the Boolean cube with different measures and adopting different sampling strategies (two sides of the same coin) can have differing effects: which leads to the current study. This paper studies the concept of weighted influences for solving MaxCon. In particular, we study endowing the Boolean cube with the Bernoulli measure and performing biased (as opposed to uniform) sampling. Theoretically, we prove the weighted influences, under this measure, of points belonging to larger structures are smaller than those of points belonging to smaller structures in general. We also consider another "natural" family of sampling/weighting strategies, sampling with uniform measure concentrated on a particular (Hamming) level of the cube. Based on weighted sampling, we modify the algorithm of Tennakoon et al., and test on both synthetic and real datasets. This paper is not promoting a new approach per se, but rather studying the issue of weighted sampling. Accordingly, we are not claiming to have produced a superior algorithm: rather we show some modest gains of Bernoulli sampling, and we illuminate some of the interactions between structure in data and weighted sampling.
Guaranteed robust nonlinear minimax estimation
IEEE Transactions on Automatic Control, 2002
Minimax parameter estimation aims at characterizing the set of all values of the parameter vector that minimize the largest absolute deviation between experimental data and corresponding model outputs. However, minimax estimation is well known to be extremely sensitive to outliers in the data resulting, e.g., of sensor failures. In this paper, a new method is proposed to robustify minimax estimation by allowing a prespecied number of absolute deviations to become arbitrarily large without modifying the estimates. By combining tools of interval analysis and constraint propagation, it becomes possible to compute the corresponding minimax estimates in an approximate but guaranteed way, even when the model output is nonlinear in its parameters. The method is illustrated on a problem where the parameters are not globally identiable, which demonstrates its ability to deal with the case where the minimax solution is not unique.
Representations of Monotone Boolean Functions by Linear Programs
2017
We introduce the notion of monotone linear-programming circuits (MLP circuits), a model of computation for partial Boolean functions. Using this model, we prove the following results. 1. MLP circuits are superpolynomially stronger than monotone Boolean circuits. 2. MLP circuits are exponentially stronger than monotone span programs. 3. MLP circuits can be used to provide monotone feasibility interpolation theorems for Lovasz-Schrijver proof systems, and for mixed Lovasz-Schrijver proof systems. 4. The Lovasz-Schrijver proof system cannot be polynomially simulated by the cutting planes proof system. This is the first result showing a separation between these two proof systems. Finally, we discuss connections between the problem of proving lower bounds on the size of MLPs and the problem of proving lower bounds on extended formulations of polytopes.
IEEE Robotics and Automation Letters
Semidefinite Programming (SDP) and Sums-of-Squares (SOS) relaxations have led to certifiably optimal non-minimal solvers for several robotics and computer vision problems. However, most non-minimal solvers rely on least squares formulations, and, as a result, are brittle against outliers. While a standard approach to regain robustness against outliers is to use robust cost functions, the latter typically introduce other non-convexities, preventing the use of existing non-minimal solvers. In this paper, we enable the simultaneous use of non-minimal solvers and robust estimation by providing a general-purpose approach for robust global estimation, which can be applied to any problem where a non-minimal solver is available for the outlier-free case. To this end, we leverage the Black-Rangarajan duality between robust estimation and outlier processes (which has been traditionally applied to early vision problems), and show that graduated nonconvexity (GNC) can be used in conjunction with non-minimal solvers to compute robust solutions, without requiring an initial guess. Although GNC's global optimality cannot be guaranteed, we demonstrate the empirical robustness of the resulting robust non-minimal solvers in applications, including point cloud and mesh registration, pose graph optimization, and image-based object pose estimation (also called shape alignment). Our solvers are robust to 70-80% of outliers, outperform RANSAC, are more accurate than specialized local solvers, and faster than specialized global solvers. We also propose the first certifiably optimal nonminimal solver for shape alignment using SOS relaxation.
Signal, Image and Video Processing, 2014
In many image processing applications, such as parametric range and motion segmentation, multiple instances of a model are fitted to data points. The most common robust fitting method, RANSAC , and its extensions are normally devised to segment the structures sequentially, treating the points belonging to other structures as outliers. Thus, the ratio of inliers is small and successful fitting requires a very large number of random samples, incurring cumbrous computation. This paper presents a new method to simultaneously fit multiple structures to data points in a single run. We model the parameters of multiple structures as a random finite set with multi-Bernoulli distribution. Simultaneous search for all structure parameters is performed by Bayesian update of the multi-Bernoulli parameters. Experiments involving segmentation of numerous structures show that our method outperforms well-known methods in terms of estimation error and computational cost. The fast convergence and high accuracy of our method make it an excellent choice for real-time estimation and segmentation of multiple structures in image processing applications.
Guaranteed non-linear estimation using constraint propagation on sets
International Journal of Control, 2001
Bounded-error estimation is the estimation of the parameter or state vector of a model from experimental data, under the assumption that some suitably dened errors should belong to some prior feasible sets. When the model outputs are linear in the vector to be estimated, a number of methods are available to enclose all estimates that are consistent with the data into simple sets such as ellipsoids, orthotopes or parallelotopes, thereby providing guaranteed set estimates. In the nonlinear case, the situation is much less developed and there are very few methods that produce such guaranteed estimates. In this paper, the discrete-time problem is cast into the more general framework of constraint satisfaction problems. Algorithms rather classical in the area of interval constraint propagation are extended by replacing intervals by more general subsets of real vector spaces. This makes it possible to propose a new algorithm that contracts the feasible domains for each uncertain variable optimally (i.e., no smaller domain could be obtained) and e¢ciently. The resulting methodology is illustrated on discrete-time nonlinear state estimation. The state at time k is estimated either from past measurement only or from all measurements assumed to be available from the start. Even in the causal case, prior information on the future value of the state and output vectors, due for instance to physical constraints, is readily taken into account.
IEEE Transactions on Automatic Control, 2021
Max Plus Linear (MPL) systems are often described by a transition function, which models the state evolution of the system, and a measurement function, which binds the measures with the system states. Methods for computing the inverse image of a point w.r.t. the measurement function are particularly interesting in applications where it is desirable to obtain informations about the system states based on the output observations. The inverse image of a set w.r.t. a nondeterministic MPL system, called uncertain MPL (uMPL) system, can be computed by using the Difference-Bound Matrices (DBM) approach. In this work we aim to use an interval analysis to propose a method to compute the inverse image of a point w.r.t. an uMPL system. The algorithm proposed has a lower worst-case complexity compared with the DBM approach as previously proposed in the literature.
Mapping Monotone Boolean Functions into Majority
IEEE Transactions on Computers, 2019
We consider the problem of decomposing monotone Boolean functions into majority-of-three operations, with a particular focus on decomposing the majority-n function. When targeting monotone Boolean functions, Shannon's expansion can be expressed by a single majority-of-three operation. We exploit this property to transform binary decision diagrams (BDDs) for monotone functions into majority-inverter graphs (MIGs), using a simple one-to-one mapping. This process highlights desirable properties for further majority graph optimization, e.g., symmetries between the inputs of primitive operations, which are not apparent from BDDs. Although our construction yields a quadratic upper bound on the number of majority-3 operations required to realize majority-n, for small n the concrete values are much smaller compared to those obtained from previous constructions which have linear and quasi-linear asymptotic upper bounds. Further, we demonstrate that minimum size MIGs, for the monotone functions majority-5 and majority-7, can be obtained applying a small number of algebraic transformations to the BDD.
Convex Maximization via Adjustable Robust Optimization
INFORMS Journal on Computing, 2022
Maximizing a convex function over convex constraints is an NP-hard problem in general. We prove that such a problem can be reformulated as an adjustable robust optimization (ARO) problem in which each adjustable variable corresponds to a unique constraint of the original problem. We use ARO techniques to obtain approximate solutions to the convex maximization problem. In order to demonstrate the complete approximation scheme, we distinguish the cases in which we have just one nonlinear constraint and multiple linear constraints. Concerning the first case, we give three examples in which one can analytically eliminate the adjustable variable and approximately solve the resulting static robust optimization problem efficiently. More specifically, we show that the norm constrained log-sum-exp (geometric) maximization problem can be approximated by (convex) exponential cone optimization techniques. Concerning the second case of multiple linear constraints, the equivalent ARO problem can ...