We propose, analyze, and test an alternating minimization algorithm for recovering images from blurry noisy observations with total variation (TV) regularization. This arises a new half-quadratic model applicable to not only the anisotropic but also isotropic forms of TV discretizations. The per-iteration computational complexity is three fast Fourier transforms. establish strong convergence properties including finite some variables relatively exponential (or q-linear in optimization...

In this paper, we propose and study the use of alternating direction algorithms for several $\ell_1$-norm minimization problems arising from sparse solution recovery in compressive sensing, including basis pursuit problem, basis-pursuit denoising both unconstrained constrained forms, as well others. We present investigate two classes derived either primal or dual forms $\ell_1$-problems. The construction consists main steps: (1) to reformulate an $\ell_1$-problem into one having partially...

We present a study of operating system errors found by automatic, static, compiler analysis applied to the Linux and OpenBSD kernels. Our approach differs from previous studies that consider manual inspection logs, testing, surveys because static is uniformly entire kernel source, though our necessarily considers less comprehensive variety than studies. In addition, automation allows us track over multiple versions source estimate how long remain in before they are fixed.We device drivers...

Recent compressive sensing results show that it is possible to accurately reconstruct certain compressible signals from relatively few linear measurements via solving nonsmooth convex optimization problems. In this paper, we propose the use of alternating direction method - a classic approach for problems with separable variables (D. Gabay and B. Mercier, ¿A dual algorithm solution nonlinear variational finite-element approximations,¿ Computer Mathematics Applications, vol. 2, pp. 17-40,...

The nuclear norm is widely used to induce low-rank solutions for many optimization problems with matrix variables. Recently, it has been shown that the augmented Lagrangian method (ALM) and alternating direction (ADM) are very efficient convex programming arising from various applications, provided resulting subproblems sufficiently simple have closed-form solutions. In this paper, we interested in application of ALM ADM some involved minimization problems. When do not solutions, propose...

Today's systems produce a rapidly exploding amount of data, and the data further derives more forming complex propagation network that we call data's lineage. There are many reasons users want to forget certain including its From privacy perspective, who become concerned with new risks system often their security if an attacker pollutes anomaly detector by injecting manually crafted into training set, must injected regain security. usability user can remove noise incorrect entries so...

Compressive sensing encodes a signal into relatively small number of incoherent linear measurements. In theory, the optimal incoherence is achieved by completely random measurement matrices. However, such matrices are often difficult and costly to implement in hardware realizations. Random Toeplitz circulant can be easily (or even naturally) realized various applications. This paper introduces fast algorithms for reconstructing signals from incomplete Computational results presented show...

The alternating direction method of multipliers (ADMM) is a popular and efficient first-order that has recently found numerous applications, the proximal ADMM an important variant it. main contributions this paper are proposition analysis class inertial ADMMs, which unify basic ideas point ADMM, for linearly constrained separable convex optimization. This methods nature because at each iteration applied to extrapolated current iterate in last movement. proposed primal-dual algorithm [A....

Multithreaded programs are hard to get right. A key reason is that the contract between developers and runtimes grants exponentially many schedules runtimes. We present Parrot, a simple, practical runtime with new developers. By default, it orders thread synchronizations in well-defined round-robin order, vastly reducing provide determinism (more precisely, deterministic synchronizations) stability (i.e., robustness against input or code perturbations, more useful property than determinism)....

In this paper, we first propose a general inertial proximal point algorithm (PPA) for the mixed variational inequality (VI) problem. Based on our knowledge, without stronger assumptions, convergence rate result is not known in literature type PPAs. Under certain conditions, are able to establish global and nonasymptotic $O(1/k)$ (under measure) of proposed PPA. We then show that both linearized augmented Lagrangian method (ALM) alternating direction multipliers (ADMM) structured convex...

Storage systems such as file systems, databases, and RAID have a simple, basic contract: you give them data, they do not lose or corrupt it. Often store the only copy, making its irrevocable loss almost arbitrarily bad. Unfortunately, their code is exceptionally hard to get right, since it must correctly recover from any crash at program point, no matter how state was smeared across volatile persistent memory.This paper describes EXPLODE, system that makes easy systematically check real...

Deterministic multithreading (DMT) eliminates many pernicious software problems caused by nondeterminism. It works constraining a program to repeat the same thread interleavings, or schedules, when given input. Despite much recent research, it remains an open challenge build both deterministic and efficient DMT systems for general programs on commodity hardware. To deterministically resolve data race, system must enforce schedule of shared memory accesses, mem-schedule, which can incur...

Image inpainting in wavelet domains refers to the recovery of an image from incomplete and/or inaccurate coefficients. To reconstruct image, total variation (TV) models have been widely used literature, and they produce high-quality reconstructed images. In this paper, we consider unconstrained, TV-regularized, $\ell_2$-data-fitting model recover image. The is solved by alternating direction method (ADM). At each iteration, ADM needs solve three subproblems, all which closed-form solutions....

For comprehensive analysis of all executable code, and fast turn-around time for transformations, it is essential to operate directly on binaries enable profiling, security hardening, architectural adaptation. Disassembling difficult, prior work relies a process virtual machine translate references the fly or inefficient binary code patching. Our Egalito recompiler leverages metadata present in current stripped x86_64 ARM64 generate complete disassembly, allows arbitrary modifications that...

Total variation (TV) regularization is popular in imagereconstruction due to its edge-preserving property. In this paper,we extend the alternating minimization algorithm recently proposedin [37] case of recovering images from randomprojections. Specifically, we propose solve TV regularizedleast squares problem by algorithms basedon classical quadratic penalty technique and alternatingminimization augmented Lagrangian function. The per-iterationcost proposed dominated two...

Accurate and robust disassembly of stripped binaries is challenging.The root the difficulty that highlevel structures, such as instruction function boundaries, are absent in must be recovered based on incomplete information.Current approaches rely heuristics or simple pattern matching to approximate recovery, but these methods often inaccurate brittle, especially across different compiler optimizations.We present XDA, a transfer-learning-based framework learns contextual dependencies machine...

We consider the covariance selection problem where variables are clustered into groups and inverse matrix is expected to have a blockwise sparse structure. This realized via penalizing maximum likelihood estimation of by group Lasso regularization. propose solve resulting log-determinant optimization with classical proximal point algorithm (PPA). At each iteration, as it difficult update primal directly, we first dual subproblem an inexact semismooth Newton-CG method then explicit formulas...

We consider image deblurring problem in the presence of impulsive noise. It is known that total variation (TV) regularization with L1-norm penalized data fitting (TVL1 for short) works reasonably well only when level noise relatively low. For high noise, TVL1 poorly. The reason all data, both corrupted and free, are equally fitting, leading to insurmountable difficulty balancing fitting. In this paper, we propose combine TV nonconvex smoothly clipped absolute deviation (SCAD) penalty (TVSCAD...

The golden ratio primal-dual algorithm (GRPDA) is a new variant of the classical Arrow--Hurwicz method for solving structured convex optimization problems, in which objective function consists sum two closed proper functions, one involves composition with linear transform. same as and popular (PDA) Chambolle Pock, GRPDA full-splitting sense that it does not rely on any subproblems or system equations iteratively. Compared PDA, an important feature permits larger primal dual stepsizes....

In this paper, we propose AdaBB, an adaptive gradient method based on the Barzilai-Borwein stepsize. The algorithm is line-search-free and parameter-free, it essentially provides a convergent variant of for general convex optimization problems. We analyze ergodic convergence objective function value iterates solving Compared with existing works along line research, our gives best lower bounds stepsize average stepsizes. Furthermore, present extensions proposed locally strongly composite...