A model of checkpoint behavior for applications that have I/O

Due to the increase and complexity of computer systems, reducing the overhead of fault tolerance techniques has become important in recent years. One technique in fault tolerance is checkpointing, which saves a snapshot with the information that has been computed up to a specific moment, suspending the execution of the application, consuming I/O resources and network bandwidth. Characterizing the files that are generated when performing the checkpoint of a parallel application is useful to determine the resources consumed and their impact on the I/O system. It is also important to characterize the application that performs checkpoints, and one of these characteristics is whether the application does I/O. In this paper, we present a model of checkpoint behavior for parallel applications that performs I/O; this depends on the application and on other factors such as the number of processes, the mapping of processes and the type of I/O used. These characteristics will also influence scalability, the resources consumed and their impact on the IO system. Our model describes the behavior of the checkpoint size based on the characteristics of the system and the type (or model) of I/O used, such as the number I/O aggregator processes, the buffering size utilized by the two-phase I/O optimization technique and components of collective file I/O operations. The BT benchmark and FLASH I/O are analyzed under different configurations of aggregator processes and buffer size to explain our approach. The model can be useful when selecting what type of checkpoint configuration is more appropriate according to the applications’ characteristics and resources available. Thus, the user will be able to know how much storage space the checkpoint consumes and how much the application consumes, in order to establish policies that help improve the distribution of resources.

     

Experience report on the application of genetic algorithms to reduce costs of the software validation process in the automotive sector during an engine control unit project

Software Quality Journal - The number of electronic control units (ECU) installed in vehicles is increasingly high. Manufacturers must improve the software quality and reduce cost by proposing...     

Adversarial example detection for DNN models: a review and experimental comparison

Deep learning (DL) has shown great success in many human-related tasks, which has led to its adoption in many computer vision based applications, such as security surveillance systems, autonomous vehicles and healthcare. Such safety-critical applications have to draw their path to success deployment once they have the capability to overcome safety-critical challenges. Among these challenges are the defense against or/and the detection of the adversarial examples (AEs). Adversaries can carefully craft small, often imperceptible, noise called perturbations to be added to the clean image to generate the AE. The aim of AE is to fool the DL model which makes it a potential risk for DL applications. Many test-time evasion attacks and countermeasures, i.e., defense or detection methods, are proposed in the literature. Moreover, few reviews and surveys were published and theoretically showed the taxonomy of the threats and the countermeasure methods with little focus in AE detection methods. In this paper, we focus on image classification task and attempt to provide a survey for detection methods of test-time evasion attacks on neural network classifiers. A detailed discussion for such methods is provided with experimental results for eight state-of-the-art detectors under different scenarios on four datasets. We also provide potential challenges and future perspectives for this research direction.

     

An industrial portrait background removal solution based on knowledge infusion

Background removal of an identity (ID) picture consists in separating the foreground (face, body, hair and clothes) from the background of the image. It is a necessary groundwork for all modern identity documents that also has many benefits for improving ID security. State of the art image processing techniques encountered several segmentation issues and offer only partial solutions. It is due to the presence of erratic components like hairs, poor contrast, luminosity variation, shadow, color overlap between clothes and background. In this paper, a knowledge infused approach is proposed that hybridizes smart image processing tasks and prior knowledge. The research is based on a divide and conquer strategy aiming at simulating the sequential attention of human when performing a manual segmentation. Knowledge is infused by considering the spatial relation between anatomic elements of the ID image (face feature, forehead, body and hair) as well as their “signal properties”. The process consists in first determining a convex hull around the person’s body including all the foreground while keeping very close to the contour between the background and the foreground. Then, a body map generated from biometric analysis associated to an automatic grab cut process is applied to reach a finer segmentation. Finally, a heuristic-based post-processing step consisting in correcting potential hair and fine boundary issues leads to the final segmentation. Experimental results show that the newly proposed architecture achieves better performances than tested current state-of-the-art methodologies including active contours, generalist popular deep learning techniques, and also two other ones considered as the smartest for portrait segmentation. This new technology has been adopted by an international company as its industrial ID foreground solution.

     

Analysis of the influence of binder properties on the mechanical response of bituminous mixtures

Asphalt mixtures are composed by a mass of aggregates (more than 90% of their total weight), which are bonded by a bituminous binder. Despite the fact that the binder is not the main component of these materials (around 5% of their total weight), it exerts a high influence on their mechanical response. In this sense, the service life of asphalt pavements will directly depend on the type of binder used, and thus an adequate choice is crucial to construct more durable roads. Because of this fact, it is necessary to know the characteristics of the bitumen in order to reduce the impact of different distresses that appear on roads. For this purpose, this paper studies the influence of the binder properties in the appearance of the main distresses that affect asphalt pavements around the world (stripping, fatigue cracking and plastic deformations). Five bitumens with different properties have been analysed during this research using diverse binder (UCL, multiple stress creep and recovery test and dynamic shear rheometer time sweep) and mixture (water sensitivity, wheel tracking and UGR-FACT) tests. The results obtained show that the properties of the binder influence the long-term performance of bituminous mixtures. In this sense, it can be said that flexible binders which are able to recover plastic deformations could extend the service life of the pavements.     

Nano-Scratching resistance of high-chromium white cast iron and its correlation with wear of cBN tool in machining

In this paper, a nano-scratch testing approach was used to measure and evaluate the abrasion wear resistance of high-chromium white cast irons in order to understand the wear mechanism in the interaction between the high-chromium white cast iron and the cBN cutting tool during the machining process. Scratch testing was performed on a nanoindentation instrument using a diamond indenter as the scratch tool. Linear multi-pass scratches in the same path were made on pre-worn surfaces of test materials. The correlation of the scratching resistance and tool wear measured from the machining is presented by the flank wear and maximum scratch depth. The appearance of the cutting edge on a cBN tool suggests that the abrasion wear is mainly related with a combined effect of the carbides and the matrix during machining the high-chromium white cast iron.     

Fast in situ 3D nanoimaging: a new tool for dynamic characterization in materials science

1. Download high-res image (269KB)
2. Download full-size image

     

Generalization of the ordinary state-based peridynamic model for isotropic linear viscoelasticity

This paper presents a generalization of the original ordinary state-based peridynamic model for isotropic linear viscoelasticity. The viscoelastic material response is represented using the thermodynamically acceptable Prony series approach. It can feature as many Prony terms as required and accounts for viscoelastic spherical and deviatoric components. The model was derived from an equivalence between peridynamic viscoelastic parameters and those appearing in classical continuum mechanics, by equating the free energy densities expressed in both frameworks. The model was simplified to a uni-dimensional expression and implemented to simulate a creep-recovery test. This implementation was finally validated by comparing peridynamic predictions to those predicted from classical continuum mechanics. An exact correspondence between peridynamics and the classical continuum approach was shown when the peridynamic horizon becomes small, meaning peridynamics tends toward classical continuum mechanics. This work provides a clear and direct means to researchers dealing with viscoelastic phenomena to tackle their problem within the peridynamic framework.     

Bond calibration method for Young’s modulus determination in the discrete element method framework

A new methodology for the calibration of bond microparameters in rocks represented by a package of joined random spherical particles in the discrete element method (DEM) framework is presented. Typically, calibration is achieved through a trial-and-error procedure using several DEM simulations of uniaxial compressive tests (UCTs). The bond calibration model (BCM) does not need a time-dependent UCT-DEM simulation to establish the relation between the microproperties of the bond and the macroproperties of the rock specimen. The BCM uses matrices to describe the interaction forces exerted by bonds and, by means of an assembly process similar to the finite element method, it can describe the complex network of bonds, enabling the model to capture small variations in particle size and bond distribution as demonstrated in this work. In this work, the BCM is presented and compared with UCT simulations performed using Esys Particle software. Multiple simulations are done with constant bond properties and different particle size ratios (\(D_{MAX}/D_{MIN})\) that cause small variations in the specimen’s Young’s modulus; these variations are well captured by the BCM with an error of <10%.