Small Local Patches Research Articles

Hippodrome: Data Race Repair Using Static Analysis Summaries

Implementing bug-free concurrent programs is a challenging task in modern software development. State-of-the-art static analyses find hundreds of concurrency bugs in production code, scaling to large codebases. Yet, fixing these bugs in constantly changing codebases represents a daunting effort for programmers, particularly because a fix in the concurrent code can introduce other bugs in a subtle way.In this work, we show how to harness compositional static analysis for concurrency bug detection, to enable a new Automated Program Repair (APR) technique for data races in large concurrent Java codebases. The key innovation of our work is an algorithm that translates procedure summaries inferred by the analysis tool for the purpose of bug reporting into small local patches that fix concurrency bugs (without introducing new ones). This synergy makes it possible to extend the virtues of compositional static concurrency analysis to APR, making our approacheffective(it can detect and fix many more bugs than existing tools for data race repair),scalable(it takes seconds to analyze and suggest fixes for sizeable codebases), andusable(generally, it does not require annotations from the users and can performcontinuousautomated repair). Our study, conducted on popular open-source projects, has confirmed that our tool automatically produces concurrency fixes similar to those proposed by the developers in the past.

Predicting high-resolution turbulence details in space and time

Predicting the fine and intricate details of a turbulent flow field in both space and time from a coarse input remains a major challenge despite the availability of modern machine learning tools. In this paper, we present a simple and effective dictionary-based approach to spatio-temporal upsampling of fluid simulation. We demonstrate that our neural network approach can reproduce the visual complexity of turbulent flows from spatially and temporally coarse velocity fields even when using a generic training set. Moreover, since our method generates finer spatial and/or temporal details through embarrassingly-parallel upsampling of small local patches, it can efficiently predict high-resolution turbulence details across a variety of grid resolutions. As a consequence, our method offers a whole range of applications varying from fluid flow upsampling to fluid data compression. We demonstrate the efficiency and generalizability of our method for synthesizing turbulent flows on a series of complex examples, highlighting dramatically better results in spatio-temporal upsampling and flow data compression than existing methods as assessed by both qualitative and quantitative comparisons.

Spatial‐spectral feature extraction of hyperspectral images using tensor‐based collaborative graph analysis

Although the collaborative graph-based discriminant analysis (CGDA) method has shown promising performance for the feature extraction of the hyperspectral image (HSI), both the intrinsic local subspace structures and spatial structural information are ignored in CGDA. To address these problems, a novel spatial-spectral feature extraction method, i.e. tensor-based collaborative graph analysis, is proposed in this letter. Specifically, the spectral similarity is utilized to calculate the Tikhonov matrix, which can constrain the testing samples to be represented by similar training samples in the collaborative representation model. To fully exploit the 3D spatial-spectral structural information, the collaborative representation model is extended to tensor space by using the third-order tensor representation of HSI, in which samples are constructed by small local patches around the central pixels. Depending on the above techniques, the quality of the collaborative graph constructed by the coefficient matrix can be significantly improved to obtain the discriminative low-dimensional features. Experiments on three HSI data sets demonstrate the superiority of the proposed method.

Two‐dimensional dictionary learning for suppressing random seismic noise

ABSTRACTDictionary learning is a successful method for random seismic noise attenuation that has been proven by some scholars. Dictionary learning–based techniques aim to learn a set of common bases called dictionaries from given noised seismic data. Then, the denoising process will be performed by assuming a sparse representation on each small local patch of the seismic data over the learned dictionary. The local patches that are extracted from the seismic section are essentially two‐dimensional matrices. However, for the sake of simplicity, almost all of the existing dictionary learning methods just convert each two‐dimensional patch into a one‐dimensional vector. In doing this, the geometric structure information of the raw data will be revealed, leading to low capability in the reconstruction of seismic structures, such as faults and dip events. In this paper, we propose a two‐dimensional dictionary learning method for the seismic denoising problem. Unlike other dictionary learning–based methods, the proposed method represents the two‐dimensional patches directly to avoid the conversion process, and thus reserves the important structure information for a better reconstruction. Our method first learns a two‐dimensional dictionary from the noisy seismic patches. Then, we use the two‐dimensional dictionary to sparsely represent all of the noisy two‐dimensional patches to obtain clean patches. Finally, the clean patches are patched back to generate a denoised seismic section. The proposed method is compared with the other three denoising methods, including FX‐decon, curvelet and one‐dimensional learning method. The results demonstrate that our method has better denoising performance in terms of signal‐to‐noise ratio, fault and amplitude preservation.

A divide-and-conquer strategy for facial landmark detection using dual-task CNN architecture

In this paper, we propose a novel deep learning-based framework for facial landmark detection. This framework takes as input face image returned by a face detector (Faster R-CNN) and generates as output a set of landmarks positions. Prior CNN-based methods often select randomly small local patches to predict an initial guess of landmarks locations. One issue with these local patches is that the adjacent landmarks might share the same regions due to the overlapping, thus, they might not convey precise information of each individual landmark. By contrast, our approach formulates this problem as a divide-conquer search for facial patches using CNN architecture in a hierarchy, where the input face image is recursively split into two cohesive non-overlapped subparts until each one contains only the region around the expected landmark. To attain better division of face topology, the search is carried out in a structured coarse-to-fine manner, where a learned hierarchical model of the face defining the granularity of each division level is introduced. We also propose a cascaded regressor to detect and refine the position of the individual landmark in each predicted non-overlapped patch. We adopt a carefully designed shallow CNN architecture so that to improve real-time performance. In addition, unlike previous cascaded methods, our regressor does not require auxiliary input such as initial landmarks locations. Extensive experiments on several challenging datasets (including MTFL, AFW, AFLW, COFW, 300W, and 300VW) show that our approach is particularly impressive in the unconstrained scenarios where it outperforms prior arts in both accuracy and efficiency.

Low-rank tensor learning for classification of hyperspectral image with limited labeled samples

Previous studies have demonstrated that integrating spatial information can potentially provide significant improvements for classification of hyperspectral image (HSI). However, it remains a challenging task to classify the high-dimensional HSI with limited number of training samples. In this paper, we propose a spectral-spatial classification framework based on low-rank tensor learning (lrTL). Unlike the traditional vector/matrix-based methods, the proposed lrTL method aims at improving the classification performance by naturally treating the HSI as a third-order tensor under the umbrella of multilinear algebra. First, small local patches containing the training (or test) samples are extracted from the original HSI cube by superpixel segmentation to preserve the structural information. Second, the lrTL algorithm is proposed to present the local patch of each test sample as a linear combination of all of the training patches. Low-rank constraint is enforced on the parameter tensor to capture the global structure of the HSI. Finally, the class label of the test sample can be determined by the minimal residual between the local patch containing the test sample and its approximation from different class subdictionaries. Experimental results on three benchmark HSI datasets demonstrate the effectiveness of the lrTL in improving the classification performance especially with limited training samples.

Hyperspectral Dimensionality Reduction by Tensor Sparse and Low-Rank Graph-Based Discriminant Analysis

Recently, sparse and low-rank graph-based discriminant analysis (SLGDA) has yielded satisfactory results in hyperspectral image (HSI) dimensionality reduction (DR), for which sparsity and low-rankness are simultaneously imposed to capture both local and global structure of hyperspectral data. However, SLGDA fails to exploit the spatial information. To address this problem, a tensor sparse and low-rank graph-based discriminant analysis (TSLGDA) is proposed in this paper. By regarding the hyperspectral data cube as a third-order tensor, small local patches centered at the training samples are extracted for the TSLGDA framework to maintain the structural information, resulting in a more discriminative graph. Subsequently, dimensionality reduction is performed on the tensorial training and testing samples to reduce data redundancy. Experimental results of three real-world hyperspectral datasets demonstrate that the proposed TSLGDA algorithm greatly improves the classification performance in the low-dimensional space when compared to state-of-the-art DR methods.

Tensor Block-Sparsity Based Representation for Spectral-Spatial Hyperspectral Image Classification

Recently, sparse representation has yielded successful results in hyperspectral image (HSI) classification. In the sparse representation-based classifiers (SRCs), a more discriminative representation that preserves the spectral-spatial information can be exploited by treating the HSI as a whole entity. Based on this observation, a tensor block-sparsity based representation method is proposed for spectral-spatial classification of HSI in this paper. Unlike traditional vector/matrix-based SRCs, the proposed method consists of tensor block-sparsity based dictionary learning and class-dependent block sparse representation. By naturally regarding the HSI cube as a third-order tensor, small local patches centered at the training samples are extracted from the HSI to maintain the structural information. All the patches are then partitioned into a number of groups, on which a dictionary learning model is constructed with a tensor block-sparsity constraint. A test sample is also expressed as a small local patch and the block sparse representation is then performed in a class-wise manner to take advantage of the class label information. Finally, the category of the test sample is determined by using the minimal residual. Experimental results of two real-world HSIs show that our proposed method greatly improves the classification performance of SRC.

TIDALLY DRIVEN DYNAMOS IN A ROTATING SPHERE

Large-scale planetary or stellar magnetic fields generated by a dynamo effect are mostly attributed to flows forced by buoyancy forces in electrically conducting fluid layers. However, these large-scale fields may also be controlled by tides, as previously suggested for the star $\tau$-boo, Mars or the Early Moon. By simulating a small local patch of a rotating fluid, \cite{Barker2014} have recently shown that tides can drive small-scale dynamos by exciting a hydrodynamic instability, the so-called elliptical (or tidal) instability. By performing global magnetohydrodynamic simulations of a rotating spherical fluid body, we investigate if this instability can also drive the observed large-scale magnetic fields. We are thus interested by the dynamo threshold and the generated magnetic field in order to test if such a mechanism is relevant for planets and stars. Rather than solving the problem in a geometry deformed by tides, we consider a spherical fluid body and add a body force to mimic the tidal deformation in the bulk of the fluid. This allows us to use an efficient spectral code to solve the magnetohydrodynamic problem. We first compare the hydrodynamic results with theoretical asymptotic results, and numerical results obtained in a truely deformed ellipsoid, which confirms the presence of the elliptical instability. We then perform magnetohydrodynamic simulations, and investigate the dynamo capability of the flow. Kinematic and self-consistent dynamos are finally simulated, showing that the elliptical instability is capable of generating dipole dominated large-scale magnetic field in global simulations of a fluid rotating sphere.

Fitting parameters of stochastic birth–death models to metapopulation data

Populations that are structured into small local patches are a common feature of ecological and epidemiological systems. Models describing this structure are often referred to as metapopulation models in ecology or household models in epidemiology. Small local populations are subject to demographic stochasticity. Theoretical studies of household disease models without resistant stages (SIS models) have shown that local stochasticity can be ignored for between patch disease transmission if the number of connected patches is large. In that case the distribution of the number of infected individuals per household reaches a stationary distribution described by a birth–death process with a constant immigration term. Here we show how this result, in conjunction with the balancing condition for birth–death processes, provides a framework to estimate demographic parameters from a frequency distribution of local population sizes. The parameter estimation framework is applicable to estimate parameters of disease transmission models as well as metapopulation models.

The long-range saliency of edge- and corner-based salient points

A major goal of salient-point (SP) detection is increasing computational efficiency. Therefore, methods which can detect saliency of a large region by evaluation of only a small local patch are of particular interest. This paper checks for well-known detectors whether saliency outreaches the actual SPs.

A Local One‐Zone Model of MagnetoHydroDynamic Turbulence in Dwarf Nova Disks

The evolution of the magnetorotational instability (MRI) during the transition from outburst to quiescence in a dwarf nova disk is investigated using three-dimensional MHD simulations. The shearing box approximation is adopted for the analysis so that the efficiency of angular momentum transport is studied in a small local patch of the disk: this is usually referred as to a one-zone model. To take account of the low ionization fraction of the disk, the induction equation includes both ohmic dissipation and the Hall effect. We induce a transition from outburst to quiescence by an instantaneous decrease of the temperature. The evolution of the MRI during the transition is found to be very sensitive to the temperature of the quiescent disk. As long as the temperature is higher than a critical value of about 2000 K, MHD turbulence and angular momentum transport is sustained by the MRI. However, MHD turbulence dies away within an orbital time if the temperature falls below this critical value. In this case, the stress drops off by more than 2 orders of magnitude and is dominated by the Reynolds stress associated with the remnant motions from the outburst. The critical temperature depends slightly on the distance from the central star and the local density of the disk.

Global Three-Dimensional Surface Reconstruction from Occluding Contours

This paper considers the problem of 3D surface reconstruction of nonpolyhedral objects through a series of images taken by a moving camera. We propose a new direct method of surface reconstruction which is based on 3D regularized uniform bicubic B-spline surface patches. This reconstruction is achieved by observing the motion of occluding contours where the view lines graze the surface. It is shown that such a reconstruction of 3D surface is possible when the motion of the camera is known, with the exception of those fully concave parts where no view line can be tangential to the surface. Once a set of overlapping B-spline surface patches which constitute the 3D surface of an object are recontructed, our developed method can fuse them efficiently to obtain a global surface. This is based on 3D triangular mesh and regularized uniform bicubic B-spline surface interpolation. This method differs from some of the previous works: First, it considers the problem of regularization on the 3D surface instead of smoothing the contours in the 2D image. Second, we propose a global surface recovery from small local patches. Previous authors have been content only with the recovery of some local properties of the surface such as 3D shape position and surface curvature estimation. Experimental results are presented for both synthetic and real data.