Fast Likelihood Computation in Speech Recognition using Matrices

Acoustic modeling using mixtures of multivariate Gaussians is the prevalent approach for many speech processing problems. Computing likelihoods against a large set of Gaussians is required as a part of many speech processing systems and it is the computationally dominant phase for Large Vocabulary Continuous Speech Recognition (LVCSR) systems. We express the likelihood computation as a multiplication of matrices representing augmented feature vectors and Gaussian parameters. The computational gain of this approach over traditional methods is by exploiting the structure of these matrices and efficient implementation of their multiplication. In particular, we explore direct low-rank approximation of the Gaussian parameter matrix and indirect derivation of low-rank factors of the Gaussian parameter matrix by optimum approximation of the likelihood matrix. We show that both the methods lead to similar speedups but the latter leads to far lesser impact on the recognition accuracy. Experiments on 1,138 work vocabulary RM1 task and 6,224 word vocabulary TIMIT task using Sphinx 3.7 system show that, for a typical case the matrix multiplication based approach leads to overall speedup of 46 % on RM1 task and 115 % for TIMIT task. Our low-rank approximation methods provide a way for trading off recognition accuracy for a further increase in computational performance extending overall speedups up to 61 % for RM1 and 119 % for TIMIT for an increase of word error rate (WER) from 3.2 to 3.5 % for RM1 and for no increase in WER for TIMIT. We also express pairwise Euclidean distance computation phase in Dynamic Time Warping (DTW) in terms of matrix multiplication leading to saving of approximately of computational operations. In our experiments using efficient implementation of matrix multiplication, this leads to a speedup of 5.6 in computing the pairwise Euclidean distances and overall speedup up to 3.25 for DTW.

Synthesis, characterization, EPR, photo and thermoluminescence properties of YAlO3:Ni2+ nanophosphors

YAlO3:Ni2+ (0.1 mol%) doped nanophosphor was synthesised by a low temperature solution combustion method. Powder X-ray diffraction (PXRD) confirms the orthorhombic phase of yttrium aluminate (YAlO3) along with traces of Y3Al5O12. Scanning Electron microscopy (SEM) shows that the powder particles appears to be spherical in shape with large agglomeration. The average crystallite sizes appeared to be in the range 45-90 nm and the same was confirmed by transmission electron microscopy (TEM) and Williamson-Hall (W-H) plots. Electron Paramagnetic Resonance (EPR) and photoluminescence (PL) studies reveal that Ni2+ ions are in octahedral coordination. Thermoluminescence (TL) glow curve consists of two peaks with the main peak at similar to 224 degrees C and a shouldered peak at 285 degrees C was recorded in the range 0.2-15 kGy gamma-irradiated samples. The TL intensity was found to be increasing linearly for 224 degrees C and 285 degrees C peaks up to 1 kGy and thereafter it shows sub-linear (up to 8 kGy) and saturation behavior. The trap parameters namely activation energy (E), order of kinetics (b), frequency factor (s) at different gamma-doses were determined using Chens glow peak shape and Luschiks methods then the results are discussed in detail. Simple glow peak structure, the 224 degrees C peak in YAlO3:Ni2+ nanophosphor can be used in personal dosimetry. (C) 2012 Elsevier B.V. All rights reserved.

SEP-optimal adaptive gain and transmit power amplify-and-forward relaying

Amplify-and-forward (AF) relay based cooperation has been investigated in the literature given its simplicity and practicality. Two models for AF, namely, fixed gain and fixed power relaying, have been extensively studied. In fixed gain relaying, the relay gain is fixed but its transmit power varies as a function of the source-relay (SR) channel gain. In fixed power relaying, the relay's instantaneous transmit power is fixed, but its gain varies. We propose a general AF cooperation model in which an average transmit power constrained relay jointly adapts its gain and transmit power as a function of the channel gains. We derive the optimal AF gain policy that minimizes the fading- averaged symbol error probability (SEP) of MPSK and present insightful and tractable lower and upper bounds for it. We then analyze the SEP of the optimal policy. Our results show that the optimal scheme is up to 39.7% and 47.5% more energy-efficient than fixed power relaying and fixed gain relaying, respectively. Further, the weaker the direct source-destination link, the greater are the energy-efficiency gains.

Estimating the fraction of progeny virions that must incorporate APOBEC3G for suppression of productive HIV-1 infection

The contest between the host factor APOBEC3G (A3G) and the HIV-1 protein Vif presents an attractive target of intervention. The extent to which the A3G-Vif interaction must be suppressed to tilt the balance in favor of A3G remains unknown. We employed stochastic simulations and mathematical modeling of the within-host dynamics and evolution of HIV-1 to estimate the fraction of progeny virions that must incorporate A3G to render productive infection unsustainable. Using three different approaches, we found consistently that a transition from sustained infection to suppression of productive infection occurred when the latter fraction exceeded similar to 0.8. The transition was triggered by A3G-induced hypermutations that led to premature stop codons compromising viral production and was consistent with driving the basic reproductive number, R-o, below unity. The fraction identified may serve as a quantitative guideline for strategies targeting the A3G-Vif axis. (C) 2013 Elsevier Inc. All rights reserved.

Isolation and characterization of biofumigant from leaves of Lantana camara for control of stored grain insect pests

Due to environmental concerns, health hazards to man and the evolution of resistance in insect pests, there have been constant efforts to discover newer insecticides both from natural sources and by chemical synthesis. Natural sources for novel molecules hold promise in view of their eco-friendly nature, selectivity and mammalian safety. We have isolated one natural bioactive molecule from the leaves of Lantana camara named Coumaran, based on various physical-chemical and spectroscopic techniques (IR, H-1 NMR, C-13 NMR and MS). Coumaran is highly toxic and very low concentration is needed for control of stored product insects. This molecule has potent grain protectant potential and caused significant reduction in F1 progeny of all the three species in the treated grain and the progeny was completely suppressed at 30 mu g/l. The differences in germination between the control and treated grains were not significant. The lack of any adverse effect of Coumaran on the seed germination is highly desirable for a grain protectant, becoming a potential source of biofumigant for economical and environmentally friendly pest control strategies against stored grain pests during storage of grains or pulses. (C) 2013 Elsevier B.V. All rights reserved.

Evaluation. of Transverse Piezoelectric Coefficient of ZnO Thin Films Deposited on Different Flexible Substrates: A Comparative Study on the Vibration Sensing Performance

We report on the systematic comparative study of highly c-axis oriented and crystalline piezoelectric ZnO thin films deposited on four different flexible substrates for vibration sensing application. The flexible substrates employed for present experimental study were namely a metal alloy (Phynox), metal (aluminum), polyimide (Kapton), and polyester (Mylar). ZnO thin films were deposited by an RF reactive magnetron sputtering technique. ZnO thin films of similar thicknesses of 700 +/- 30 nm were deposited on four different flexible substrates to have proper comparative studies. The crystallinity, surface morphology, chemical composition, and roughness of ZnO thin films were evaluated by respective material characterization techniques. The transverse piezoelectric coefficient (d(31)) value for assessing the piezoelectric property of ZnO thin films on different flexible substrates was measured by a four-point bending method. ZnO thin films deposited on Phynox alloy substrate showed relatively better material characterization results and a higher piezoelectric d(31) coefficient value as compared to ZnO films on metal and polymer substrates. In order to experimentally verify the above observations, vibration sensing studies were performed. As expected, the ZnO thin film deposited on Phynox alloy substrate showed better vibration sensing performance. It has generated the highest peak to peak output voltage amplitude of 256 mV as compared to that of aluminum (224 mV), Kapton (144 mV), and Mylar (46 mV). Therefore, metal alloy flexible substrate proves to be a more suitable, advantageous, and versatile choice for integrating ZnO thin films as compared to metal and polymer flexible substrates for vibration sensing applications. The present experimental study is extremely important and helpful for the selection of a suitable flexible substrate for various applications in the field of sensor and actuator technology.

Synthesis of benzimidazole-grafted graphene oxide/multi-walled carbon nanotubes composite for supercapacitance application

We are reporting the fabrication, characterizations and supercapacitance performance of benzimidazole-grafted graphene oxide/multi-walled carbon nanotubes (BI-GO/MWCNTs) composite. The synthesis of BI-GO materials involves cyclization reaction of carboxylic groups on GO among the hydroxyl and amino groups on o-phenylenediamine. The BI-GO/MWCNTs composite has been fabricated via in situ reduction of BI-GO using hydrazine in presence of MWCNTs. Scanning electron microscopy (SEM), Transmission electron microscopy (TEM), Raman spectroscopy, X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR) have been used to characterize its surface and elemental composition. The uniform dispersion of MWCNTs with BI-GO helps to improve the charge transfer reaction during electrochemical process. The specific capacitance of BI-GO/MWCNTs composite is 275 and 460 F/g at 200 and 5 mV/s scan rate in 1 mol/L aqueous solution of H2SO4. This BI-GO/MWCNTs composite has shown 224 F/g capacitance after 1300 cycles at 200 mV/s scan rate, which represents its good electrochemical stability. (C) 2014 Elsevier B.V. All rights reserved.

Controlling air solubility to maintain `'Cassie'' state for sustained drag reduction

`'Cassie'' state of wetting can be established by trapping air pockets on the crevices of textured hydrophobic surfaces, leading to significant drag reduction. However, this drag reduction cannot be sustained due to gradual dissolution of trapped air into water. In this paper, we explore the possibility of sustaining the underwater Cassie state of wetting in a microchannel by controlling the solubility of air in water; the solubility being changed by controlling the local absolute pressure near the surface. We show that using this method, we can in fact make the water locally supersaturated with air thus encouraging the growth of trapped air pockets on the surface. In this case, the water acts as a pumping medium, delivering air to the crevices of the hydrophobic surface in the microchannel, where the presence of air pockets is most beneficial from the drag reduction perspective. In our experiments, the air trapped on a textured surface is visualized using total internal reflection based technique, at different local absolute pressures with the pressure drop (or drag) also being simultaneously measured. We find that, by controlling the pressure and hence the solubility close to the surface, we can either shrink or grow the trapped air bubbles, uniformly over a large surface area. The experiments show that, by precisely controlling the pressure and hence the solubility we can sustain the `'Cassie state'' over extended periods of time. This method thus provides a means of getting sustained drag reduction from a textured hydrophobic surface in channel flows. (C) 2014 Elsevier B.V. All rights reserved.

PriDyn: Enabling Differentiated I/O Services in Cloud Using Dynamic Priorities

Virtualization is one of the key enabling technologies for Cloud computing. Although it facilitates improved utilization of resources, virtualization can lead to performance degradation due to the sharing of physical resources like CPU, memory, network interfaces, disk controllers, etc. Multi-tenancy can cause highly unpredictable performance for concurrent I/O applications running inside virtual machines that share local disk storage in Cloud. Disk I/O requests in a typical Cloud setup may have varied requirements in terms of latency and throughput as they arise from a range of heterogeneous applications having diverse performance goals. This necessitates providing differential performance services to different I/O applications. In this paper, we present PriDyn, a novel scheduling framework which is designed to consider I/O performance metrics of applications such as acceptable latency and convert them to an appropriate priority value for disk access based on the current system state. This framework aims to provide differentiated I/O service to various applications and ensures predictable performance for critical applications in multi-tenant Cloud environment. We demonstrate through experimental validations on real world I/O traces that this framework achieves appreciable enhancements in I/O performance, indicating that this approach is a promising step towards enabling QoS guarantees on Cloud storage.

Chemical-free graphene by unzipping carbon nanotubes using cryo-milling

Current paper reports synthesis of chemical free graphene by unzipping of the carbon nanotubes (CNTs) using high strain rate deformation at 150K. A specially designed cryomill operating at 150 K was used for the experiments. The mechanism of unzipping was further explored using molecular dynamics (MD) simulations. Both experimental and simulation results reveal two modes of unzipping through radial and shear loading. (C) 2015 Elsevier Ltd. All rights reserved.

Synthesis of Instruction Extensions on HyperCell, a Reconfigurable Datapath

In this paper we present HyperCell as a reconfigurable datapath for Instruction Extensions (IEs). HyperCell comprises an array of compute units laid over a switch network. We present an IE synthesis methodology that enables post-silicon realization of IE datapaths on HyperCell. The synthesis methodology optimally exploits hardware resources in HyperCell to enable software pipelined execution of IEs. Exploitation of temporal reuse of data in HyperCell results in significant reduction of input/output bandwidth requirements of HyperCell.

An enzyme-linked immuno focus assay for rapid detection and enumeration, and a newborn mouse model for human non-polio enteroviruses associated with acute diarrhea

We have recently reported significant association of non-polio enteroviruses (NPEVs) with acute and persistent diarrhea (18-21% of total diarrheal cases), and non-diarrheal Increased Frequency of Bowel Movements (IFoBM-ND) (about 29% of the NPEV infections) in children and that the NPEV-associated diarrhea was as significant as rotavirus diarrhea. However, their diarrhea-causing potential is yet to be demonstrated in an animal model system. Since the determination of virus titers by the traditional plaque assay takes 4-7 days, there is a need for development of a rapid method for virus titer determination to facilitate active clinical research on enterovirus-associated diarrhea. The goal of this study is to develop a cell-based rapid detection and enumeration method and to demonstrate the diarrhea-inducing potential of purified and characterized non-polio enteroviruses, which were isolated from diarrheic children. Here we describe generation of monoclonal and polyclonal antibodies against purified strains belonging to different serotypes, and development of an enzyme-linked immuno focus assay (ELIFA) for detection and enumeration of live NPEV particles in clinical and purified virus samples, and a newborn mouse model for NPEV diarrhea. Plaque-purified NPVEs, belonging to different serotypes, isolated from children with diarrhea, were grown in cell culture and purified by isopycnic CsCl density gradient centrifugation. By ELIFA, NPEVs could be detected and enumerated within 12 h post-infection. Our results demonstrated that Coxsackievirus B1 (CVB1) and CVB5 strains, isolated from diarrheic children, induced severe diarrhea in orally-inoculated 9-12 day-old mouse pups, fulfilling Koch's postulates. The methods described here would facilitate studies on NPEV-associated gastrointestinal disease. (C) 2015 Elsevier B.V. All rights reserved.

Systematic Analysis of Mycobacterial Acylation Reveals First Example of Acylation-mediated Regulation of Enzyme Activity of a Bacterial Phosphatase

Protein lysine acetylation is known to regulate multiple aspects of bacterial metabolism. However, its presence in mycobacterial signal transduction and virulence-associated proteins has not been studied. In this study, analysis of mycobacterial proteins from different cellular fractions indicated dynamic and widespread occurrence of lysine acetylation. Mycobacterium tuberculosis proteins regulating diverse physiological processes were then selected and expressed in the surrogate host Mycobacterium smegmatis. The purified proteins were analyzed for the presence of lysine acetylation, leading to the identification of 24 acetylated proteins. In addition, novel lysine succinylation and propionylation events were found to co-occur with acetylation on several proteins. Protein-tyrosine phosphatase B (PtpB), a secretory phosphatase that regulates phosphorylation of host proteins and plays a critical role in Mycobacterium infection, is modified by acetylation and succinylation at Lys-224. This residue is situated in a lid region that covers the enzyme's active site. Consequently, acetylation and succinylation negatively regulate the activity of PtpB.

Studies on Carbon Mediated Paste Screen Printed Sensors for Blood Glucose Sensing Application

Ready-to-use screen printed glucose sensors are fabricated using Prussian Blue (PB) and Cobalt Phthalocyanine (CoPC) mediated carbon inks as working electrodes. The reference and counter electrodes are screen printed using silver/silver chloride and graphitic carbon paste respectively. The screen printed reference electrodes (internal reference electrode (IRE)) are found to be stable for more than 60 minutes when examined with saturated calomel electrode. Optimal operating voltage for PB and CoPC screen printed sensors are determined by hydrodynamic voltammetric technique. Glucose oxidase is immobilized on the working electrodes by cross-linking method. PB mediated glucose sensor exhibits a sensitivity of 5.60 mA cm(-2)/mM for the range, 10 to 1000 mu M. Sensitivity of CoPC mediated glucose sensor is found to be 5.224 mu A cm(-2)/mM and amperometeric response is linear for the range, 100 to 1500 mu M. Interference studies on the fabricated glucose sensors are conducted with species like uric acid and ascorbic acid. PB mediated sensors showed a completely interference-free behavior. The sensing characteristics of PB mediated glucose sensors are also studied in diluted human serum samples and the results are compared with the values obtained through standard clinical method. The co-efficient of variation is found to be less than 5%. (C) 2015 The Electrochemical Society. All rights reserved.