Showing papers by "fondazione bruno kessler published in 2011"

ADJUST: An automatic EEG artifact detector based on the joint use of spatial and temporal features.

Abstract: A successful method for removing artifacts from electroencephalogram (EEG) recordings is Independent Component Analysis (ICA), but its implementation remains largely user-dependent. Here, we propose a completely automatic algorithm (ADJUST) that identifies artifacted independent components by combining stereotyped artifact-specific spatial and temporal features. Features were optimized to capture blinks, eye movements, and generic discontinuities on a feature selection dataset. Validation on a totally different EEG dataset shows that (1) ADJUST’s classification of independent components largely matches a manual one by experts (agreement on 95.2% of the data variance), and (2) Removal of the artifacted components detected by ADJUST leads to neat reconstruction of visual and auditory eventrelated potentials from heavily artifacted data. These results demonstrate that ADJUST provides a fast, efficient, and automatic way to use ICA for artifact removal. Descriptors: Electroencephalography, Independent component analysis, EEG artifacts, EEG artefacts, Event-related potentials, Ongoing brain activity, Automatic classification, Thresholding

Sensitivity studies for third-generation gravitational wave observatories

Abstract: Advanced gravitational wave detectors, currently under construction, are expected to directly observe gravitational wave signals of astrophysical origin. The Einstein Telescope (ET), a third-generation gravitational wave detector, has been proposed in order to fully open up the emerging field of gravitational wave astronomy. In this paper we describe sensitivity models for ET and investigate potential limits imposed by fundamental noise sources. A special focus is set on evaluating the frequency band below 10 Hz where a complex mixture of seismic, gravity gradient, suspension thermal and radiation pressure noise dominates. We develop the most accurate sensitivity model, referred to as ET-D, for a third-generation detector so far, including the most relevant fundamental noise contributions.

Complete electric dipole response and the neutron skin in 208Pb.

Abstract: A benchmark experiment on Pb-208 shows that polarized proton inelastic scattering at very forward angles including 0 degrees is a powerful tool for high-resolution studies of electric dipole (E1) and spin magnetic dipole (M1) modes in nuclei over a broad excitation energy range to test up-to-date nuclear models. The extracted E1 polarizability leads to a neutron skin thickness r(skin) = 0.156(-0.021)(+0.025) fm in Pb-208 derived within a mean-field model [Phys. Rev. C 81, 051303 (2010)], thereby constraining the symmetry energy and its density dependence relevant to the description of neutron stars.

A 160×128 single-photon image sensor with on-pixel 55ps 10b time-to-digital converter

Abstract: Image sensors capable of resolving the time-of-arrival (ToA) of individual photons with high resolution are needed in several applications, such as fluorescence lifetime imaging microscopy (FLIM), Forster resonance energy transfer (FRET), optical rangefinding, and positron emission tomography In FRET, for example, typical fluorescence lifetime is of the order of 100 to 300ps, thus deep-subnanosecond resolutions are needed in the instrument response function (IRF) This in turn requires new time-resolved image sensors with better time resolution, increased throughput, and lower costs Solid-state avalanche photodiodes operated in Geiger-mode, or single-photon avalanche diodes (SPADs), have existed for decades [1] but only recently have SPADs been integrated in CMOS However, as array sizes have grown, the readout bottleneck has also become evident, leading to hybrid designs or more integration and more parallelism on-chip [2,3] This trend has accelerated with the introduction of SPAD devices in deep-submicron CMOS, that have enabled the design of massively parallel arrays where the entire photon detection and ToA circuitry is integrated on-pixel [4,5]

Three-dimensional imaging of chemical phase transformations at the nanoscale with full-field transmission X-ray microscopy

Abstract: The ability to probe morphology and phase distribution in complex systems at multiple length scales unravels the interplay of nano- and micrometer-scale factors at the origin of macroscopic behavior. While different electron- and X-ray-based imaging techniques can be combined with spectroscopy at high resolutions, owing to experimental time limitations the resulting fields of view are too small to be representative of a composite sample. Here a new X-ray imaging set-up is proposed, combining full-field transmission X-ray microscopy (TXM) with X-ray absorption near-edge structure (XANES) spectroscopy to follow two-dimensional and three-dimensional morphological and chemical changes in large volumes at high resolution (tens of nanometers). TXM XANES imaging offers chemical speciation at the nanoscale in thick samples (>20 µm) with minimal preparation requirements. Further, its high throughput allows the analysis of large areas (up to millimeters) in minutes to a few hours. Proof of concept is provided using battery electrodes, although its versatility will lead to impact in a number of diverse research fields.

Safety, Dependability and Performance Analysis of Extended AADL Models

Abstract: This paper presents a component-based modelling approach to system-software co-engineering of real-time embedded systems, in particular aerospace systems. Our method is centred around the standardized Architecture Analysis and Design Language (AADL) modelling framework. We formalize a significant subset of AADL, incorporating its recent Error Model Annex for modelling faults and repairs. The major distinguishing aspects of this component-based approach are the possibility to describe nominal hardware and software operations, hybrid (and timing) aspects, as well as probabilistic faults and their propagation and recovery. Moreover, it supports dynamic (i.e. on-the-fly) reconfiguration of components and inter-component connections. The operational semantics gives a precise interpretation of specifications by providing a mapping onto networks of event-data automata. These networks are then subject to different kinds of formal analysis such as model checking, safety and dependability analysis and performance evaluation. Mature tool support realizes these analyses. The activities reported in this paper are carried out in the context of the correctness, modelling, and performance of aerospace systems, project which is funded by the European Space Agency.

A Dual-band Millimeter-wave Kinetic Inductance Camera for the IRAM 30 m Telescope

Abstract: The Neel IRAM KIDs Array (NIKA) is a fully integrated measurement system based on kinetic inductance detectors (KIDs) currently being developed for millimeter wave astronomy. The instrument includes dual-band optics allowing simultaneous imaging at 150 GHz and 220 GHz. The imaging sensors consist of two spatially separated arrays of KIDs. The first array, mounted on the 150 GHz branch, is composed of 144 lumped-element KIDs. The second array (220 GHz) consists of 256 antenna-coupled KIDs. Each of the arrays is sensitive to a single polarization; the band splitting is achieved by using a grid polarizer. The optics and sensors are mounted in a custom dilution cryostat, with an operating temperature of ~70 mK. Electronic readout is realized using frequency multiplexing and a transmission line geometry consisting of a coaxial cable connected in series with the sensor array and a low-noise 4 K amplifier. The dual-band NIKA was successfully tested in 2010 October at the Institute for Millimetric Radio Astronomy (IRAM) 30 m telescope at Pico Veleta, Spain, performing in-line with laboratory predictions. An optical NEP was then calculated to be around 2 × 10–16 W Hz–1/2 (at 1 Hz) while under a background loading of approximately 4 pW pixel–1. This improvement in comparison with a preliminary run (2009) verifies that NIKA is approaching the target sensitivity for photon-noise limited ground-based detectors. Taking advantage of the larger arrays and increased sensitivity, a number of scientifically relevant faint and extended objects were then imaged including the Galactic Center SgrB2 (FIR1), the radio galaxy Cygnus A, and the NGC1068 Seyfert galaxy. These targets were all observed simultaneously in the 150 GHz and 220 GHz atmospheric windows.

Video-rate fluorescence lifetime imaging camera with CMOS single-photon avalanche diode arrays and high-speed imaging algorithm

Abstract: A high-speed and hardware-only algorithm using a center of mass method has been proposed for single-detector fluorescence lifetime sensing applications. This algorithm is now implemented on a field programmable gate array to provide fast lifetime estimates from a 32 × 32 low dark count 0.13 ?m complementary metaloxide-semiconductor single-photon avalanche diode (SPAD) plus time-to-digital converter array. A simple look-up table is included to enhance the lifetime resolvability range and photon economics, making it comparable to the commonly used least-square method and maximum likelihood estimation based software. To demonstrate its performance, a widefield microscope was adapted to accommodate the SPAD array and image different test samples. Fluorescence lifetime imaging microscopy on fluorescent beads in Rhodamine 6G at a frame rate of 50 fps is also shown.

Tactile-Data Classification of Contact Materials Using Computational Intelligence

Abstract: The two major components of a robotic tactile-sensing system are the tactile-sensing hardware at the lower level and the computational/software tools at the higher level. Focusing on the latter, this research assesses the suitability of computational-intelligence (CI) tools for tactile-data processing. In this context, this paper addresses the classification of sensed object material from the recorded raw tactile data. For this purpose, three CI paradigms, namely, the support-vector machine (SVM), regularized least square (RLS), and regularized extreme learning machine (RELM), have been employed, and their performance is compared for the said task. The comparative analysis shows that SVM provides the best tradeoff between classification accuracy and computational complexity of the classification algorithm. Experimental results indicate that the CI tools are effective in dealing with the challenging problem of material classification.

Symbolic search-based testing

Abstract: We present an algorithm for constructing fitness functions that improve the efficiency of search-based testing when trying to generate branch adequate test data. The algorithm combines symbolic information with dynamic analysis and has two key advantages: It does not require any change in the underlying test data generation technique and it avoids many problems traditionally associated with symbolic execution, in particular the presence of loops. We have evaluated the algorithm on industrial closed source and open source systems using both local and global search-based testing techniques, demonstrating that both are statistically significantly more efficient using our approach. The test for significance was done using a one-sided, paired Wilcoxon signed rank test. On average, the local search requires 23.41% and the global search 7.78% fewer fitness evaluations when using a symbolic execution based fitness function generated by the algorithm.

Analog circuit for timing measurements with large area SiPMs coupled to LYSO crystals

Abstract: The most common method for time pick-off from signals coming from SiPMs coupled to scintillator crystals in PET applications is the Leading Edge Triggering. In this work we propose a new filtering scheme to be applied before the discriminator. It implements a type of baseline compensation aimed at the reduction of the time jitter due to the dark counts of the detector and is based on a simple analog filter, which is well suited for an ASIC implementation. We describe a circuit, built with discrete components, that implements the filter. Finally we report on the Coincidence Resolving Time measurements performed coupling the circuit to the real detector, composed of a SiPM built at FBK and a LYSO crystal, for the detection of 511 keV photons. The results obtained are promising, showing that the method is quite effective in reducing the impact of the detector noise on the timing performance of the system.

Convolutive BSS of Short Mixtures by ICA Recursively Regularized Across Frequencies

Abstract: This paper proposes a new method of frequency-domain blind source separation (FD-BSS), able to separate acoustic sources in challenging conditions. In frequency-domain BSS, the time-domain signals are transformed into time-frequency series and the separation is generally performed by applying independent component analysis (ICA) at each frequency envelope. When short signals are observed and long demixing filters are required, the number of time observations for each frequency is limited and the variance of the ICA estimator increases due to the intrinsic statistical bias. Furthermore, common methods used to solve the permutation problem fail, especially with sources recorded under highly reverberant conditions. We propose a recursively regularized implementation of the ICA (RR-ICA) that overcomes the mentioned problem by exploiting two types of deterministic knowledge: 1) continuity of the demixing matrix across frequencies; 2) continuity of the time-activity of the sources. The recursive regularization propagates the statistics of the sources across frequencies reducing the effect of statistical bias and the occurrence of permutations. Experimental results on real-data show that the algorithm can successfully perform a fast separation of short signals (e.g., 0.5-1s), by estimating long demixing filters to deal with highly reverberant environments (e.g., ms).

A Range Image Sensor Based on 10- $\mu{\hbox {m}}$ Lock-In Pixels in 0.18- $\mu$ m CMOS Imaging Technology

Abstract: This paper presents the design and characterization of a lock-in pixel array based on a buried channel photo-detector aimed at time-of-flight range imaging. The proposed photo-demodulator has been integrated in a 10-μm pixel pitch with a fill factor of 24%, and is capable of a maximum demodulation frequency of 50 MHz with a contrast of 29.5%. The sensor has been fabricated in a 0.18-μm CMOS imaging technology and assembled in a range camera system setup. The system provides a stream of three-dimensional images at 5-20 fps on a 3-6 m range, with a linearity error lower than 0.7% and a repeatability of 5-16 cm, while the best achievable precision is 2.7 cm at a 50-MHz modulation frequency.

Assessment of NER solutions against the first and second CALBC Silver Standard Corpus.

Abstract: Background: Competitions in text mining have been used to measure the performance of automatic text processing solutions against a manually annotated gold standard corpus (GSC). The preparation of the GSC is time-consuming and costly and the final corpus consists at the most of a few thousand documents annotated with a limited set of semantic groups. To overcome these shortcomings, the CALBC project partners (PPs) have produced a large-scale annotated biomedical corpus with four different semantic groups through the harmonisation of annotations from automatic text mining solutions, the first version of the Silver Standard Corpus (SSC-I). The four semantic groups are chemical entities and drugs (CHED), genes and proteins (PRGE), diseases and disorders (DISO) and species (SPE). This corpus has been used for the First CALBC Challenge asking the participants to annotate the corpus with their text processing solutions. Results: All four PPs from the CALBC project and in addition, 12 challenge participants (CPs) contributed annotated data sets for an evaluation against the SSC-I. CPs could ignore the training data and deliver the annotations from their genuine annotation system, or could train a machine-learning approach on the provided preannotated data. In general, the performances of the annotation solutions were lower for entities from the categories CHED and PRGE in comparison to the identification of entities categorized as DISO and SPE. The best performance over all semantic groups were achieved from two annotation solutions that have been trained on the SSC-I. The data sets from participants were used to generate the harmonised Silver Standard Corpus II (SSC-II), if the participant did not make use of the annotated data set from the SSC-I for training purposes. The performances of the participants’ solutions were again measured against the SSC-II. The performances of the annotation solutions showed again better results for DISO and SPE in comparison to CHED and PRGE.

Geometric investigation of a gaming active device

Abstract: 3D imaging systems are widely available and used for surveying, modeling and entertainment applications, but clear statements regarding their characteristics, performances and limitations are still missing. The VDI/VDE and the ASTME57 committees are trying to set some standards but the commercial market is not reacting properly. Since many new users are approaching these 3D recording methodologies, clear statements and information clarifying if a package or system satisfies certain requirements before investing are fundamental for those users who are not really familiar with these technologies. Recently small and portable consumer-grade active sensors came on the market, like TOF rangeimaging cameras or low-cost triangulation-based range sensor. A quite interesting active system was produced by PrimeSense and launched on the market thanks to the Microsoft Xbox project with the name of Kinect. The article reports the geometric investigation of the Kinect active sensors, considering its measurement performances, the accuracy of the retrieved range data and the possibility to use it for 3D modeling application.

Studies of silicon photomultipliers at cryogenic temperatures

Abstract: We investigate the behavior of silicon photomultipliers (SiPMs) at low temperatures: I–V characteristics, breakdown voltage, dark noise, afterpulsing, crosstalk, pulse shape, gain and photon detection efficiency are studied as a function of temperature in the range 50 K T 320 K . We discuss our measurements on the basis of the temperature dependent properties of silicon and of the models related to carrier generation, transport and multiplication in high electric field. We conclude that SiPMs provide an excellent alternative to vacuum tube photomultipliers (PMTs) in low temperature environments, even better than in room temperature ones: in particular they excel in the interval 100 K T 200 K .

KRATOS: a software model checker for SystemC

Abstract: The growing popularity of SystemC has attracted research aimed at the formal verification of SystemC designs. In this paper we present KRATOS, a software model checker for SystemC. KRATOS verifies safety properties, in the form of program assertions, by allowing users to explore two directions in the verification. First, by relying on the translation from SystemC designs to sequential C programs, KRATOS is capable of model checking the resulting C programs using the symbolic lazy predicate abstraction technique. Second, KRATOS implements a novel algorithm, called ESST, that combines Explicit state techniques to deal with the SystemC Scheduler, with Symbolic techniques to deal with the Threads. KRATOS is built on top of NUSMV and MATHSAT, and uses state-ofthe-art SMT-based techniques for program abstractions and refinements.

Exploring Text Virality in Social Networks

Abstract: This paper aims to shed some light on the concept of virality - especially in social networks - and to provide new insights on its structure. We argue that: (a) virality is a phenomenon strictly connected to the nature of the content being spread, rather than to the influencers who spread it (b) virality is a phenomenon with many facets, i.e. under this generic term several different effects of persuasive communication are comprised and they only partially overlap. To give ground to our claims, we provide initial experiments in a machine learning framework to show how various aspects of virality can be independently predicted according to content features.

Multi-layered monitoring and adaptation

Abstract: Service-based applications have become more and more multi-layered in nature, as we tend to build software as a service on top of infrastructure as a service. Most existing SOA monitoring and adaptation techniques address layer-specific issues. These techniques, if used in isolation, cannot deal with real-world domains, where changes in one layer often affect other layers, and information from multiple layers is essential in truly understanding problems and in developing comprehensive solutions. In this paper we propose a framework that integrates layer specific monitoring and adaptation techniques, and enables multi-layered control loops in service-based systems. The proposed approach is evaluated on a medical imaging procedure for Computed Tomography (CT) Scans, an e-Health scenario characterized by strong dependencies between the software layer and infrastructural resources.

The Automatic Recognition of Emotions in Speech

Abstract: In this chapter, we focus on the automatic recognition of emotional states using acoustic and linguistic parameters as features and classifiers as tools to predict the ‘correct’ emotional states. We first sketch history and state of the art in this field; then we describe the process of ‘corpus engineering’, i.e. the design and the recording of databases, the annotation of emotional states, and further processing such as manual or automatic segmentation. Next, we present an overview of acoustic and linguistic features that are extracted automatically or manually. In the section on classifiers, we deal with topics such as the curse of dimensionality and the sparse data problem, classifiers, and evaluation. At the end of each section, we point out important aspects that should be taken into account for the planning or the assessment of studies. The subject area of this chapter is not emotions in some narrow sense but in a wider sense encompassing emotion-related states such as moods, attitudes, or interpersonal stances as well. We do not aim at an in-depth treatise of some specific aspects or algorithms but at an overview of approaches and strategies that have been used or should be used.

Shallow Discourse Parsing with Conditional Random Fields

Abstract: Parsing discourse is a challenging natural language processing task. In this paper we take a data driven approach to identify arguments of explicit discourse connectives. In contrast to previous work we do not make any assumptions on the span of arguments and consider parsing as a token-level sequence labeling task. We design the argument segmentation task as a cascade of decisions based on conditional random fields (CRFs). We train the CRFs on lexical, syntactic and semantic features extracted from the Penn Discourse Treebank and evaluate feature combinations on the commonly used test split. We show that the best combination of features includes syntactic and semantic features. The comparative error analysis investigates the performance variability over connective types and argument positions.

A 160 $\times$ 120-Pixels Range Camera With In-Pixel Correlated Double Sampling and Fixed-Pattern Noise Correction

Abstract: This paper presents the design and electro-optical test of a 160 × 120-pixels CMOS sensor specifically conceived for Time-Of-Flight 3D imaging. The in-pixel processing allows the implementation of Indirect Time-Of-Flight technique for distance measurement with reset noise removal through Correlated Double Sampling and embedded fixed-pattern noise reduction, whereas a fast readout operation allows the pixels values to be streamed out at a maximum rate of 10 MSample/s. The imager can operate as a fast 2D camera up to 458 fps, as a 3D camera up to 80 fps, or even coupling both operation modes. The chip has been fabricated using a standard 0.18 μm 1P4M 1.8 V CMOS technology with MIM capacitors. The resulting pixel has a pitch of 29.1 μm with a fill-factor of 34% and includes 66 transistors. Distance measurements up to 4.5 m have been performed with pulsed laser light, achieving a best precision of 10 cm at 1 m in real-time at 55 fps and 175 mA current consumption.

The Transition to a Giant Vortex Phase in a Fast Rotating Bose-Einstein Condensate

Abstract: We study the Gross-Pitaevskii (GP) energy functional for a fast rotating Bose-Einstein condensate on the unit disc in two dimensions. Writing the coupling parameter as 1/e2 we consider the asymptotic regime e → 0 with the angular velocity Ω proportional to (e2|log e|)−1. We prove that if Ω = Ω0(e2|log e|)−1 and Ω0 > 2(3π)−1 then a minimizer of the GP energy functional has no zeros in an annulus at the boundary of the disc that contains the bulk of the mass. The vorticity resides in a complementary ‘hole’ around the center where the density is vanishingly small. Moreover, we prove a lower bound to the ground state energy that matches, up to small errors, the upper bound obtained from an optimal giant vortex trial function, and also that the winding number of a GP minimizer around the disc is in accord with the phase of this trial function.

Low-noise single Photon Avalanche Diodes in 0.15 μm CMOS technology

Abstract: Two different Single-Photon Avalanche Diode (SPAD) structures in a standard 0.15-nm CMOS technology are presented. A characterization of the two detectors, having a 10-μm active-area diameter, and monolithically integrated with a passive quenching circuit and a fast comparator is presented. The two devices exhibit respectively a typical dark count rate of 230cps and 160cps, an afterpulsing probability of 2.1% and 1.3% at 30ns dead time, a Photon Detection Probability of 31% and 26 % at λ=470nm and a timing resolution of 170ps and 60ps. The adopted technology is therefore promising for the realization of SPAD-based image sensors with good overall performance.

High energy-resolution zero-degree facility for light-ion scattering and reactions at iThemba LABS

Abstract: The setup and experimental techniques for measurements of zero-degree inelastic scattering and reactions involving light ions with the K=600 magnetic spectrometer at iThemba LABS are described. Measurements were performed for inelastic proton scattering at an incident energy of 200 MeV for targets ranging from 27Al to 208Pb. An energy-resolution of 45 keV (FWHM) was achieved by utilizing the faint-beam dispersion-matching technique. A background subtraction procedure was applied and allowed for the extraction of excitation energy spectra with low background. Measurements of the (p,t) reaction at zero degrees for Ep=100 and 200 MeV benefited from the difference in magnetic rigidity between the reaction products and the beam particles, resulting in background-free spectra with an excitation energy-resolution of 32 and 48 keV (FWHM), respectively, and a scattering angle resolution of 0.55° (FWHM). The addition of Double Sided Silicon Strip Detectors (DSSSD) at backward scattering angles allowed for coincident measurements of particle-decay of states excited in the (p,t) reaction at E p = 200 MeV .

Earth mover's prototypes: A convex learning approach for discovering activity patterns in dynamic scenes

Abstract: We present a novel approach for automatically discovering spatio-temporal patterns in complex dynamic scenes Similarly to recent non-object centric methods, we use low level visual cues to detect atomic activities and then construct clip histograms Differently from previous works, we formulate the task of discovering high level activity patterns as a prototype learning problem where the correlation among atomic activities is explicitly taken into account when grouping clip histograms Interestingly at the core of our approach there is a convex optimization problem which allows us to efficiently extract patterns at multiple levels of detail The effectiveness of our method is demonstrated on publicly available datasets