An optimisation of Gaussian mixture models for integer processing units

This paper investigates sub-integer implementations of the adaptive Gaussian mixture model (GMM) for background/foreground segmentation to allow the deployment of the method on low cost/low power processors that lack Floating Point Unit (FPU). We propose two novel integer computer arithmetic techniques to update Gaussian parameters. Specifically, the mean value and the variance of each Gaussian are updated by a redefined and generalised "round'' operation that emulates the original updating rules for a large set of learning rates. Weights are represented by counters that are updated following stochastic rules to allow a wider range of learning rates and the weight trend is approximated by a line or a staircase. We demonstrate that the memory footprint and computational cost of GMM are significantly reduced, without significantly affecting the performance of background/foreground segmentation.

Energetic beams of negative and neutral hydrogen from intense laser plasma interaction

We present observations of intense beams of energetic negative hydrogen ions and fast neutral hydrogen atoms in intense (5 × 10 W/cm) laser plasma interaction experiments, which were quantified in numerical calculations. Generation of negative ions and neutral atoms is ascribed to the processes of electron capture and loss by a laser accelerated positive ion in the collisions with a cloud of droplets. A comparison with a numerical model of charge exchange processes provides information on the cross section of the electron capture in the high energy domain.

Table-Top Laser-Based Source of Femtosecond, Collimated, Ultrarelativistic Positron Beams

The generation of ultrarelativistic positron beams with short duration (τ_e⁺≃30 fs), small divergence (θ_e⁺≃3 mrad), and high density (n _e⁺≃1014-1015 cm^-3) from a fully optical setup is reported. The detected positron beam propagates with a high-density electron beam and γ rays of similar spectral shape and peak energy, thus closely resembling the structure of an astrophysical leptonic jet. It is envisaged that this experimental evidence, besides the intrinsic relevance to laser-driven particle acceleration, may open the pathway for the small-scale study of astrophysical leptonic jets in the laboratory. 

Laser-driven generation of collimated ultra-relativistic positron beams

We report on recent experimental results concerning the generation of collimated (divergence of the order of a few mrad) ultra-relativistic positron beams using a fully optical system. The positron beams are generated exploiting a quantum-electrodynamic cascade initiated by the propagation of a laser-accelerated, ultra-relativistic electron beam through high-Z solid targets. As long as the target thickness is comparable to or smaller than the radiation length of the material, the divergence of the escaping positron beam is of the order of the inverse of its Lorentz factor. For thicker solid targets the divergence is seen to gradually increase, due to the increased number of fundamental steps in the cascade, but it is still kept of the order of few tens of mrad, depending on the spectral components in the beam. This high degree of collimation will be fundamental for further injection into plasma-wakefield afterburners.

Effect of magnetic bias on Gaussian pulse scattering by stacked nonlinear semiconductor layers

The nonlinear scattering of pulses by periodic stacks of semiconductor layers with magnetic bias has been studied in the self-consistent problem formulation, taking into account mobility of carriers. The three-wave mixing technique has been applied to the analysis of the waveform evolution in the stacks illuminated by two Gaussian pulses with different central frequencies and lengths. The effects of external magnetic bias, and stack physical and geometrical parameters on the properties of the scattered waveforms are discussed. © 2013 IEEE.

Gaussian States in Quantum Information

These notes originated out of a set of lectures in Quantum Optics and Quantum Information given by one of us (MGAP) at the University of Napoli and the University of Milano. A quite broad set of issues are covered, ranging from elementary concepts to current research topics, and from fundamental concepts to applications. A special emphasis has been given to the phase space analysis of quantum dynamics and to the role of Gaussian states in continuous variable quantum information.

Instrumentation for diagnostics and control of laser-accelerated proton (ion) beams

Suitable instrumentation for laser-accelerated proton (ion) beams is critical for development of integrated, laser-driven ion accelerator systems. Instrumentation aimed at beam diagnostics and control must be applied to the driving laser pulse, the laser-plasma that forms at the target and the emergent proton (ion) bunch in a correlated way to develop these novel accelerators. This report is a brief overview of established diagnostic techniques and new developments based on material presented at the first workshop on 'Instrumentation for Diagnostics and Control of Laser-accelerated Proton (Ion) Beams' in Abingdon, UK. It includes radiochromic film (RCF), image plates (IP), micro-channel plates (MCP), Thomson spectrometers, prompt inline scintillators, time and space-resolved interferometry (TASRI) and nuclear activation schemes. Repetition-rated instrumentation requirements for target metrology are also addressed. (C) 2013 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.

Design of a compact spectrometer for high-flux MeV gamma-ray beams

A novel design for a compact gamma-ray spectrometer is presented. The proposed system allows for spectroscopy of high-flux multi-MeV gamma-ray beams with MeV energy resolution in a compact design. In its basic configuration, the spectrometer exploits conversion of gamma-rays into electrons via Compton scattering in a low-Z material. The scattered electron population is then spectrally resolved using a magnetic spectrometer. The detector is shown to be effective for gamma-ray energies between 3 and 20 MeV. The main properties of the spectrometer are confirmed by Monte Carlo simulations.

Measurements of high-energy radiation generation from laser-wakefield accelerated electron beams

Using high-energy (∼0.5 GeV) electron beams generated by laser wakefield acceleration (LWFA), bremsstrahlung radiation was created by interacting these beams with various solid targets. Secondary processes generate high-energy electrons, positrons, and neutrons, which can be measured shot-to-shot using magnetic spectrometers, short half-life activation, and Compton scattering. Presented here are proof-of-principle results from a high-resolution, high-energy gamma-ray spectrometer capable of single-shot operation, and high repetition rate activation diagnostics. We describe the techniques used in these measurements and their potential applications in diagnosing LWFA electron beams and measuring high-energy radiation from laser-plasma interactions.

A dynamic sampling methodology for plasma etch processes using Gaussian process regression

Plasma etch is a key process in modern semiconductor manufacturing facilities as it offers process simplification and yet greater dimensional tolerances compared to wet chemical etch technology. The main challenge of operating plasma etchers is to maintain a consistent etch rate spatially and temporally for a given wafer and for successive wafers processed in the same etch tool. Etch rate measurements require expensive metrology steps and therefore in general only limited sampling is performed. Furthermore, the results of measurements are not accessible in real-time, limiting the options for run-to-run control. This paper investigates a Virtual Metrology (VM) enabled Dynamic Sampling (DS) methodology as an alternative paradigm for balancing the need to reduce costly metrology with the need to measure more frequently and in a timely fashion to enable wafer-to-wafer control. Using a Gaussian Process Regression (GPR) VM model for etch rate estimation of a plasma etch process, the proposed dynamic sampling methodology is demonstrated and evaluated for a number of different predictive dynamic sampling rules. © 2013 IEEE.

Efficient Marginal Likelihood Computation for Gaussian Process Regression

In a Bayesian learning setting, the posterior distribution of a predictive model arises from a trade-off between its prior distribution and the conditional likelihood of observed data. Such distribution functions usually rely on additional hyperparameters which need to be tuned in order to achieve optimum predictive performance; this operation can be efficiently performed in an Empirical Bayes fashion by maximizing the posterior marginal likelihood of the observed data. Since the score function of this optimization problem is in general characterized by the presence of local optima, it is necessary to resort to global optimization strategies, which require a large number of function evaluations. Given that the evaluation is usually computationally intensive and badly scaled with respect to the dataset size, the maximum number of observations that can be treated simultaneously is quite limited. In this paper, we consider the case of hyperparameter tuning in Gaussian process regression. A straightforward implementation of the posterior log-likelihood for this model requires O(N^3) operations for every iteration of the optimization procedure, where N is the number of examples in the input dataset. We derive a novel set of identities that allow, after an initial overhead of O(N^3), the evaluation of the score function, as well as the Jacobian and Hessian matrices, in O(N) operations. We prove how the proposed identities, that follow from the eigendecomposition of the kernel matrix, yield a reduction of several orders of magnitude in the computation time for the hyperparameter optimization problem. Notably, the proposed solution provides computational advantages even with respect to state of the art approximations that rely on sparse kernel matrices.

Influence of micro and macro cracks due to sustained loading on chloride-induced corrosion of reinforced concrete beams

Chloride-induced corrosion of steel is one of the most commonly found problems affecting the durability of reinforced concrete structures in both marine environment and where de-icing salt is used in winter. As the significance of micro-cracks on chloride induced corrosion is not well documented, 24 reinforced concrete beams (4 different mixes - one containing Portland cement and another containing 35% ground granulated blastfurnace slag at 0.45 and 0.65 water-binder ratios) were subjected to three levels of sustained lateral loading (0%, 50% and 100% of the load that can induce 0.1 mm wide cracks on the tension surface of beam - F_0.1) in this work. The beams were then subjected to weekly cycles of wetting with 10% NaCl solution for 1 day followed by 6 days of drying at 20 (±1) °C up to an exposure period of 60 weeks. The progress of corrosion of steel was monitored using half-cell potential apparatus and linear polarisation resistance (LPR) test. These results have shown that macro-cracks (at load F_0.1) and micro-cracks (at 50% of F_0.1) greatly accelerated both the initiation and propagation stages of the corrosion of steel in the concrete beams. Lager crack widths for the F_0.1 load cases caused higher corrosion rates initially, but after about 38 weeks of exposure, there was a decrease in the rate of corrosion. However, such trends could not be found in 50% F _0.1 group of beams. The extent of chloride ingress also was influenced by the load level. These findings suggest that the effect of micro-cracking at lower loads are very important for deciding the service life of reinforced concrete structures in chloride exposure environments. © 2014 4th International Conference on the Durability of Concrete Structures.

Post-tensioning glulam timber beams with basalt FRP tendons

Improvements in the structural performance of glulam timber beams by the inclusion of reinforcing materials can increase both the service performance and ultimate capacity. This paper describes a series of four-point bending tests conducted, under service loads and to failure, on unreinforced, reinforced and post-tensioned glulam timber beams, where the reinforcing tendon used is 12 mm dia. basalt fibre-reinforced polymer. The research is designed to evaluate the benefits offered by including an active reinforcement in contrast to the passive reinforcement typically used within timber strengthening works, in addition to establishing the effect that bonding the reinforcing tendon has on the material's performance. Further experimental tests have also been developed to investigate the long-term implications of this research, with emphasis placed upon creep and loss of post-tensioning; however, this is ongoing and is not presented in this paper. The laboratory investigations establish that the flexural strength and stiffness increase for both the unbonded and bonded post-tensioned timbers compared to the unreinforced and reinforced beams. Timber that is post-tensioned with an unbonded basalt fibre-reinforced polymer tendon shows a flexural strength increase of 2ṡ8% and an increase in stiffness of 8ṡ7%. Post-tensioned beams with a bonded basalt fibre-reinforced polymer tendon show increases in flexural strength and stiffness of 15ṡ4% and 11ṡ5% respectively.