Timing isolation and improved scheduling of deep neural networks for real-time systems

In recent years, the performance of deep neural networks (DNNs) is significantly improved, making them suitable for many application fields, such as autonomous driving, advanced robotics, and industrial control. Despite a lot of research being devoted to improving the accuracy of DNNs, only limited efforts have been spent to enhance their timing predictability, required in several real-time applications. This paper proposes a software infrastructure based on the Linux operating system to integrate DNNs within a real-time multicore system. It has been realized by modifying both the internal scheduler of the popular TensorFlow framework and the SCHED_DEADLINE scheduling class of Linux. The proposed infrastructure allows providing timing isolation of DNN inference tasks, hence improving the determinism of the temporal interference generated by TensorFlow. The proposal is finally evaluated with a case study derived from a state-of-the-art benchmark inspired by an autonomous industrial system. Extensive experiments demonstrate the effectiveness of the proposed solution and show a significant reduction of both average and longest-observed response times of TensorFlow tasks.     

Grappa quality from the Chianti and Montepulciano areas (Tuscany,Italy): monitoring the leaching of copper from distillation columns

The main aim of this work was to evaluate the leaching of copper in grappa from distillation plant that treats Cabernet and Sangiovese marcs (Montepulciano and Chianti, Tuscany). Copper content (determined by atomic absorption spectrometry, AAS) was 1.05 ± 0.03 mg L^?1 in industrial distillates (Cabernet marc, 82.3% alcohol, v/v), and 3.7 ± 0.2 μg L^?1 in artificial grappa (62.0%, Cabernet) obtained by distilling marc in a laboratory‐scale glassware plant. The artificial grappa distillate was refluxed over copper shavings, and copper content, C_Cu, was analysed in relation to reflux time and wear of the copper bell. The results showed that C_Cu trend was almost linear with increasing alcohol concentration and constant reflux time, and vice versa. The ratios of selected volatile components were not influenced by reflux on the shavings.     

Economic crisis and firm exit: do intangibles matter?

ABSTRACT

The financial crisis has caused many business closures, especially in the periphery of the European Monetary Union. In this paper, we use original firm-level Italian data to explore the role of intangibles in limiting firm exit during the crisis. While intangibles strengthen firms’ resilience, i.e. the ability to cope with unexpected shocks, they also entail sunk costs, which expose firms to greater financial risk. In the longer term, when access to external finance is critical for survival, we expect intangibles to play a positive role only if combined with a solid financial structure. Our results support these hypotheses: intangibles directly reduce the probability of firm exit during the initial phase of the crisis (before 2010). At later stages, the beneficial effect of intangible assets is conditional on the firm exhibiting a solid pre-crisis financial status. Managerial and policy implications are discussed.     

Transport properties of photogenerated charge carriers in black diamond films

Surface nanostructured diamond films by ultrashort pulse laser treatment, named black diamond, demonstrated a huge increase in the photogeneration capability for photons with sub-bandgap energy (<5.47?eV). Here we analyze in detail the transport properties of photogenerated charge carriers in several black diamond samples, in order to better understand the electronic behavior of defect levels introduced by the laser treatment. If compared with the pristine diamond films, the mean charge carriers’ mobility-lifetime product, evaluated from the over bandgap photocurrent characteristics, remarkably increases in every black diamond set up to a defined absorbed accumulated laser fluence before decreasing at the highest fluence values. We attribute this effect to the laser-induced introduction of fast traps for one charge carrier type, that increases with fluence. At the same time, an increasing density of recombination centers, capturing permanently the charge carriers, is formed. A trade-off treatment condition can be found in order to maximize the sensitivity to sub-bandgap photons and the mean mobility-lifetime product as well as to limit the effect of recombination centers.     

Comprehensive numerical simulation of threshold-voltage transients in nitride memories

A new non-local algorithm for accurately calculating the band-to-band tunneling current suitable for TCAD semiconductor simulators is proposed in this paper. The proposed algorithm captures the essential physics of multi-dimensional tunneling in a 2D structure, and is designed to be robust and to achieve independence on the mesh grid. The new algorithm enables accurate modeling of T-FET and investigation of its device physics.     

Weakly Constrained Lucy–Richardson with Applications to Inversion of Light Scattering Data

Lucy–Richardson (LR) is a classical iterative regularization method largely used for the restoration of nonnegative solutions. LR finds applications in many physical problems, such as for the inversion of light scattering data. In these problems, there is often additional information on the true solution that is usually ignored by many restoration methods because the related measurable quantities are likely to be affected by non-negligible noise. In this article we propose a novel Weakly Constrained Lucy–Richardson (WCLR) method which adds a weak constraint to the classical LR by introducing a penalization term, whose strength can be varied over a very large range. The WCLR method is simple and robust as the standard LR, but offers the great advantage of widely stretching the domain range over which the solution can be reliably recovered. Some selected numerical examples prove the performances of the proposed algorithm.     

Kinetics and Isotherms of Mercury Biosorption by Dry Biomass of Arthrospira (Spirulina) platensis

Inorganic mercury was effectively removed from water solutions by means of dry biomass of Arthrospira (Spirulina) platensis as biosorbent. The results obtained at defined pH and biomass concentration as well as variable Hg concentrations were used to estimate biosorption kinetics and isotherm parameters. Biosorption kinetics was best described by the pseudo-first-order model of Lagergren that allowed estimating an average specific rate constant as high as 0.044 min⁻¹, while the Dubinin and Radushkevich model was the best fitting isotherm. Hg biosorption seemed to occur via physisorption. Considering the wide dissemination of A. platensis and its easy cultivation, these results point to its biomass as a low-cost biosorbent to treat Hg-contaminated waters.     

Unsupervised digital movie restoration with spatial models of color

In this paper, we present an efficient approach to digital color restoration, based on the idea of recovering the appearance of color rather than the original color signal, since, in most of the cases, for old films the original reference is missing, and new films or digital coding can be subject to severe gamut transformations. This approach is based on the application of algorithms inspired by the capabilities of the Human Vision System of automatically adjust to the variation of color and lightness in the scene. The proposed method allows more unsupervised restoration. We present an overview of the approach, characteristics of this family of algorithms, and restoration examples.