Extended VTS for noise-robust speech recognition

Model compensation is a standard way of improving the robustness of speech recognition systems to noise. A number of popular schemes are based on vector Taylor series (VTS) compensation, which uses a linear approximation to represent the influence of noise on the clean speech. To compensate the dynamic parameters, the continuous time approximation is often used. This approximation uses a point estimate of the gradient, which fails to take into account that dynamic coefficients are a function of a number of consecutive static coefficients. In this paper, the accuracy of dynamic parameter compensation is improved by representing the dynamic features as a linear transformation of a window of static features. A modified version of VTS compensation is applied to the distribution of the window of static features and, importantly, their correlations. These compensated distributions are then transformed to distributions over standard static and dynamic features. With this improved approximation, it is also possible to obtain full-covariance corrupted speech distributions. This addresses the correlation changes that occur in noise. The proposed scheme outperformed the standard VTS scheme by 10% to 20% relative on a range of tasks. © 2006 IEEE.

Microstructure and superconducting properties of single grains of Y-Ba-Cu-O containing Y-2411(M) and Y2O3

Y-Ba-Cu-O (YBCO) single grains have the potential to generate large trapped magnetic fields for engineering applications, and research on the processing and properties of this material has attracted interest world-wide over the past 20 years. In particular, the introduction of flux pinning centers to the large grain microstructure to improve its current density Jc, and hence trapped field, has been investigated extensively. Y2Ba 4CuMO2 [Y-2411(M)], where M = Nb, Ta, Mo, W, Ru, Zr, Bi and Ag, has been discovered recently to form very effective flux pinning centers due primarily to its ability to form nano-size inclusions in the superconducting phase matrix. However, the addition of the Y-2411(M) phase to the precursor composition complicates the melt-processing of single grains. The addition of Y2O3 to the precursor composition, however, broadens the growth window of single YBCO grains containing Y-2411 (M). We report an investigation of the microstructures and superconducting properties of single grains of this composition grown by top seeded melt growth (TSMG). © 2010 IEEE.

Nanotube-based passively mode-locked Ytterbium-pumped Raman laser

Over the past decades mode-locked fibre lasers have been extensively refined and developed, with most research efforts focussing on employing rare-earth doped fibres as the active elements [1]. This presents the problem that operation is limited to regions of the spectrum where such elements exhibit gain [1]. Raman amplification in silica fibre is an attractive way to overcome this spectral limitation, with gain available across the entire transparency window (300 nm - 2300 nm) [2-4]. There have been a number of reports utilising Raman gain in ultrashort pulse sources [2-4], however none using a broadband saturable absorber, such as carbon nanotubes [5-7] and graphene [7-9]. A broadband saturable absorber is an essential pre-requisite in order to fully exploit the wavelength flexibility provided by the Raman gain in short pulse mode-locked fiber lasers. © 2011 IEEE.

Colourless adaptively modulated optical OFDM transmitters using SOAs as intensity modulators.

The wavelength dependent transmission performance of adaptively modulated optical OFDM (AMOOFDM) signals is investigated, for the first time, over optical amplification- and chromatic dispersion compensation-free IMDD SMF systems using semiconductor optical amplifiers (SOAs) as intensity modulators. A theoretical SOA model describing both optical gain saturation and gain spectral dynamics is developed, based on which optimum SOA operating conditions are identified for various wavelengths varying in a broad range of 1510 nm- 1590 nm. Results show that, SOA intensity modulators operating at the identified optimum conditions enable the realization of colourless AMOOFDM transmitters within the aforementioned wavelength window. Such transmitters are capable of supporting >30 Gb/s signal transmission over 60 km SMFs.

External corporate venture capital investment: Towards a framework for capturing and measuring strategic value

This paper presents the initial results of on-going research in the field of external Corporate Venture Capital (CVC) investments, i.e. equity investments of large corporations in entrepreneurial ventures which originated outside the corporation. The research is motivated by the fact that external CVC plays an increasingly important role within the strategy of corporations. Driven by a general trend towards a more open approach to innovation, companies see particular value in external corporate venturing as a tool to gain, for example, access to complementary technologies and a general window on technology developments. The review of literature in the field of external corporate venturing clearly reveals that theoretical gaps exist in understanding mechanisms for capturing value and measurements of this value. To help close these gaps, the research addresses the underlying question "How do corporations and start-ups capture and measure strategic value through external CVC investments" by using embedded, multiple case studies. Following an initial set of case studies, steps towards the development of a framework for capturing and measuring strategic value from CVC investments are outlined within this paper and the resulting preliminary framework is presented. The paper closes with an outlook on ongoing and future research steps. © 2009 PICMET.

The development of a hot rolling process for variable cross-section I-beams

To meet targeted reductions in CO 2 emissions by 2050, demand for metal must be cut, for example through the use of lightweight technologies. However, the efficient production of weight optimized components often requires new, more flexible forming processes. In this paper, a novel hot rolling process is presented for forming I-beams with variable cross-section, which are lighter than prismatic alternatives. First, the new process concept is presented and described. A detailed computational and experimental analysis is then conducted into the capabilities of the process. Results show that the process is capable of producing defect free I-beams with variations in web depth of 30-50%. A full analysis of the process then indicates the likely failure modes, and identifies a safe operating window. Finally, the implications of these results for producing lightweight beams are discussed. © 2012 Elsevier B.V. All rights reserved.

Two is better than one: Distinct roles for familiarity and recollection in retrieving palimpsest memories

Storing a new pattern in a palimpsest memory system comes at the cost of interfering with the memory traces of previously stored items. Knowing the age of a pattern thus becomes critical for recalling it faithfully. This implies that there should be a tight coupling between estimates of age, as a form of familiarity, and the neural dynamics of recollection, something which current theories omit. Using a normative model of autoassociative memory, we show that a dual memory system, consisting of two interacting modules for familiarity and recollection, has best performance for both recollection and recognition. This finding provides a new window onto actively contentious psychological and neural aspects of recognition memory.

Multiple color reflection in a single unit cell using double-layer electrochromic reaction

Multiple color states have been realized in single unit cell using double electrochromic (EC) reaction. The precise control of bistability in EC compounds which can maintain several colors on the two separated electrodes allows this new type of pixel to be realized. The specific electrical driving gives a way to maintain both sides in the reduced EC states and this colors overlapping in the vertical view direction can achieve the black state. The four color states (G, B, W, BK) in one cell/pixel can make a valuable progress to achieve a high quality color devices such like electronic paper, outdoor billboard, smart window and flexible display using external light source. © 2012 Optical Society of America.

Innovative Stereo Vision-Based Approach to Generate Dense Depth Map of Transportation Infrastructure

Three-dimensional (3-D) spatial data of a transportation infrastructure contain useful information for civil engineering applications, including as-built documentation, on-site safety enhancements, and progress monitoring. Several techniques have been developed for acquiring 3-D point coordinates of infrastructure, such as laser scanning. Although the method yields accurate results, the high device costs and human effort required render the process infeasible for generic applications in the construction industry. A quick and reliable approach, which is based on the principles of stereo vision, is proposed for generating a depth map of an infrastructure. Initially, two images are captured by two similar stereo cameras at the scene of the infrastructure. A Harris feature detector is used to extract feature points from the first view, and an innovative adaptive window-matching technique is used to compute feature point correspondences in the second view. A robust algorithm computes the nonfeature point correspondences. Thus, the correspondences of all the points in the scene are obtained. After all correspondences have been obtained, the geometric principles of stereo vision are used to generate a dense depth map of the scene. The proposed algorithm has been tested on several data sets, and results illustrate its potential for stereo correspondence and depth map generation.

Innovative Stereo Vision-Based Approach to Generate Dense Depth Map of Transportation Infrastructure

Three-dimensional (3-D) spatial data of a transportation infrastructure contain useful information for civil engineering applications, including as-built documentation, on-site safety enhancements, and progress monitoring. Several techniques have been developed for acquiring 3-D point coordinates of infrastructure, such as laser scanning. Although the method yields accurate results, the high device costs and human effort required render the process infeasible for generic applications in the construction industry. A quick and reliable approach, which is based on the principles of stereo vision, is proposed for generating a depth map of an infrastructure. Initially, two images are captured by two similar stereo cameras at the scene of the infrastructure. A Harris feature detector is used to extract feature points from the first view, and an innovative adaptive window-matching technique is used to compute feature point correspondences in the second view. A robust algorithm computes the nonfeature point correspondences. Thus, the correspondences of all the points in the scene are obtained. After all correspondences have been obtained, the geometric principles of stereo vision are used to generate a dense depth map of the scene. The proposed algorithm has been tested on several data sets, and results illustrate its potential for stereo correspondence and depth map generation.

Experimental investigation of stratified combustion: Application of laser diagnostic techniques

In order to better understand the stratified combustion, the propagation of flame through stratified mixture field in laminar and turbulent flow conditions has been studied by using combined PIV/PLIF techniques. A great emphasis was placed on developing methods to improve the accuracy of local measurements of flame propagation. In particular, a new PIV approach has been developed to measure the local fresh gas velocity near preheat zone of flame front. To improve the resolution of measurement, the shape of interrogation window has been continuously modified based on the local flame topology and gas expansion effect. Statistical analysis of conditioned local measurements by the local equivalence ratio of flames allows the characterization of the properties of flame propagation subjected to the mixture stratification in laminar and turbulent flows, especially the highlight of the memory effect.

The structure of turbulent stratified and premixed methane/air flames I: Non-swirling flows

A series of flames in a turbulent methane/air stratified swirl burner is presented. The degree of stratification and swirl are systematically varied to generate a matrix of experimental conditions, allowing their separate and combined effects to be investigated. Non-swirling flows are considered in the present paper, and the effects of swirl are considered in a companion paper (Part II). A mean equivalence ratio of φ=0.75 is used, with φ for the highest level of stratification spanning 0.375-1.125. The burner features a central bluff-body to aid flame stabilization, and the influence of the induced recirculation zone is also considered. The current work focuses on non-swirling flows where two-component particle image velocimetry (PIV) measurements are sufficient to characterize the main features of the flow field. Scalar data obtained from Rayleigh/Raman/CO laser induced fluorescence (CO-LIF) line measurements at 103μm resolution allow the behavior of key combustion species-CH 4, CO 2, CO, H 2, H 2O and O 2-to be probed within the instantaneous flame front. Simultaneous cross-planar OH-PLIF is used to determine the orientation of the instantaneous flame normal in the scalar measurement window, allowing gradients in temperature and progress variable to be angle corrected to their three dimensional values. The relationship between curvature and flame thickness is investigated using the OH-PLIF images, as well as the effect of stratification on curvature.The main findings are that the behavior of the key combustion species in temperature space is well captured on the mean by laminar flame calculations regardless of the level of stratification. H 2 and CO are significant exceptions, both appearing at elevated levels in the stratified flames. Values for surface density function and by extension thermal scalar dissipation rate are found to be substantially lower than laminar values, as the thickening of the flame due to turbulence dominates the effect of increased strain. These findings hold for both premixed and stratified flames. The current series of flames is proposed as an interesting if challenging set of test cases for existing and emerging turbulent flame models, and data are available on request. © 2012 The Combustion Institute.

On optical reflection at heterojunction interface of thin film solar cells

Heterojunction is an important structure for the development of photovoltaic solar cells. In contrast to homojunction structures, heterojunction solar cells have internal crystalline interfaces, which will reflect part of the incident light, and this has not been considered carefully before though many heterostructure solar cells have been commercialized. This paper discusses the internal reflection for various material systems used for the development of heterostructure-based solar cells. It has been found that the most common heterostructure solar cells have internal reflection less than 2%, while some potential heterojunction solar cells such as ITO/GaAs, ITO/InP, Si/Ge, polymer/semiconductors and oxide semiconductors may have internal reflection as high as 20%. Also it is worse to have a window layer with a lower refractive index than the absorption layer for solar cells. Ignoring this strong internal reflection will lead to severe deterioration and reduction of conversion efficiency; therefore measures have to be taken to minimize or prevent this internal reflection. © 2013 Elsevier B.V.

Room temperature all-silicon photonic crystal nanocavity light emitting diode at sub-bandgap wavelengths

Silicon is now firmly established as a high performance photonic material. Its only weakness is the lack of a native electrically driven light emitter that operates CW at room temperature, exhibits a narrow linewidth in the technologically important 1300-1600 nm wavelength window, is small and operates with low power consumption. Here, an electrically pumped all-silicon nano light source around 1300-1600 nm range is demonstrated at room temperature. Using hydrogen plasma treatment, nano-scale optically active defects are introduced into silicon, which then feed the photonic crystal nanocavity to enhance the electrically driven emission in a device via Purcell effect. A narrow (Δλ=0.5 nm) emission line at 1515 nm wavelength with a power density of 0.4mW/cm2 is observed, which represents the highest spectral power density ever reported from any silicon emitter. A number of possible improvements are also discussed, that make this scheme a very promising light source for optical interconnects and other important silicon photonics applications. © 2012 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

Anisotropic AC Behavior of Multifilamentary Bi-2223/Ag Tapes

In this communication, we report on the anisotropy of the superconducting properties of multifilamentary Bi-based tapes experimentally investigated by AC magnetic susceptibility measurements. The susceptibility $\chi= \chi' - j \chi''$ was measured using a commercial system and a couple of orthogonal pick-up coils. The $\chi''$ vs. temperature curves were shown to exhibit two peaks. The smaller of the peaks, occurring near T = 72K, was only visible for particular field directions and within a given frequency window. Such results point out the role played by the phase difference between the applied magnetic field and the internal magnetic field seen by the filaments.