Combining formant frequency based on variable order LPC coding with acoustic features for TIMIT phone recognition

Combination of multiple acoustic features has great potential to improve Automatic Speech Recognition (ASR) accuracy. Our contribution in this research was to investigate one novel method when using voiced formants’ features in combination with standard MFCC features in order to enhance TIMIT phone recognition. These voiced features provide accurate formants frequencies using a Variable Order LPC Coding (VO-LPC) algorithm that was combined with continuity constraints. The overall estimating formants were concatenated with MFCC features when a voiced frame could be detected. For feature-level combination, Heteroscedastic Linear Discriminant Analysis (HLDA) based approach had been used successfully to find an optimal linear combination of successive vectors of a single feature stream.     

Tight bounds for the identical parallel machine scheduling problem

We address the problem of minimizing makespan on identical parallel machines. We propose new lower bounding strategies and heuristics for this fundamental scheduling problem. The lower bounds are based on the so‐called lifting procedure. In addition, two optimization‐based heuristics are proposed. These heuristics require iteratively solving a subset‐sum problem. We present the results of computational experiments that provide strong evidence that the new proposed lower and upper bounds consistently outperform the best bounds from the literature.     

Fuzzy logic-based pre-classifier for tropical wood species recognition system

Classifying tropical wood species poses a considerable economic challenge and failure to classify the wood species accurately can have significant effects on timber industries. The problem of wood recognition is compounded with the nonlinearities of the features among the similar wood species. Besides that, large wood databases presented a problem of large processing time especially for online wood recognition system. In view of these problems, we propose the use of fuzzy logic-based pre-classifier as a means of treating uncertainty to improve the classification accuracy of tropical wood recognition system. The pre-classifier serve as a clustering mechanism for the large database simplifying the classification process making it more efficient. The use of the fuzzy logic-based pre-classifier has managed to increase the accuracy of the wood recognition system by 4 % and reduce the processing time for training and testing by more than 75 % and 26 % respectively.     

Limited-complexity model-unfalsifying adaptive tracking-control

Some recently developed methods for robust tracking control by model unfalsification require demanding computations with polytopes. This paper introduces a limited-complexity polytope updating technique that can be used with adaptive control schemes. The main result establishes the convergence of adaptive steady-state tracking performance to near the ideal robust performance. This result is extended to a robust control algorithm with limited-complexity polytope updating.     

Defining and computing a value based cyber-security measure

In earlier work, we presented a value based measure of cybersecurity that quantifies the security of a system in concrete terms, specifically, in terms of how much each system stakeholder stands to lose (in dollars per hour of operation) as a result of security threats and system vulnerabilities; our metric varies according to the stakes that each stakeholder has in meeting each security requirement. In this paper, we discuss the specification and design of a system that collects, updates, and maintains all the information that pertains to estimating our cybersecurity measure, and offers stakeholders quantitative means to make security-related decisions.     

Coextrusion of superconducting wires

Cu/Y₁Ba₂Cu₃O_7?x superconducting composite wires have been fabricated using the powder-in-tube coextrusion process. Increase in load during the extrusion process seems to indicate that powder particles from the deformation zone of the die travel up toward the unextruded part of the billet and gradually increase the density in that part to some saturation value, which prohibits further extrusion at moderate to heavy applied load. A powder removal technique has been devised to perform full length extrusion under such conditions. Powder removal, however, is not required if the extrusion ratio is reduced sufficiently. The extrusion ratio that will give a full length extrusion without powder removal seems to be a function of starting billet length. Extrusion ratio of 2.52 gives smooth full length extrusions with 25.4 mm (1 in.) starting billet length at reasonable load values. Four probe resistivity and magnetic susceptibility measurements indicate a critical temperature (Tc) value of the formed wire to be close to 88 K. Density measurements using the volume displacement method indicate a value of 4.819 g/cm³, which is approximately 76.5% of the theoretical density. M.M. Dehghani and A. Ahmad are Assistant Professor and     

Building micro and nanosystems with electrochemical discharges

Since the discovery of the electrochemical discharge phenomenon by Fizeau and Foucault, several contributions have expanded the wide range of applications associated with this high current density electrochemical process. The complexity of the phenomenon, from the macroscopic to the microscopic scales, led since then to experimental and theoretical studies from different research fields. This contribution reviews the chemical and electrochemical perspectives where a mechanistic model based on results from radiation chemistry of aqueous solutions is proposed. In addition applications to micro-machining and fabrication of nanoparticles are discussed.     

Experimental investigation of particle distribution in a flow through a stenosed artery

The particle distribution of a dilute solid-liquid suspension through a stenosed arterial geometry was investigated. Particle image velocimetry (PIV) was used to determine the velocity as well as to acquire the flow images. The light intensity scattered by particles was evaluated to determine the particle distribution. Flow separation exists where the flow emerges from the stenosis throat. From the PIV images, the particle density distribution exhibited differing non-uniform characteristics which vary with flow rate, particle size and particle concentration. At low flow rates, a particle-free layer is formed. As the flow rate is increased, particles accumulate in concentric recirculation orbits within the downstream vortex. Particles with larger size and higher concentration tend to accumulate more towards the center of the vortex.     

An analytical model for the performance evaluation of stack‐based Web cache replacement algorithms

Web caching has been the solution of choice to web latency problems. The efficiency of a Web cache is strongly affected by the replacement algorithm used to decide which objects to evict once the cache is saturated. Numerous web cache replacement algorithms have appeared in the literature. Despite their diversity, a large number of them belong to a class known as stack‐based algorithms. These algorithms are evaluated mainly via trace‐driven simulation. The very few analytical models reported in the literature were targeted at one particular replacement algorithm, namely least recently used (LRU) or least frequently used (LFU). Further they provide a formula for the evaluation of the Hit Ratio only. The main contribution of this paper is an analytical model for the performance evaluation of any stack‐based web cache replacement algorithm. The model provides formulae for the prediction of the object Hit Ratio, the byte Hit Ratio, and the delay saving ratio. The model is validated against extensive discrete event trace‐driven simulations of the three popular stack‐based algorithms, LRU, LFU, and SIZE, using NLANR and DEC traces. Results show that the analytical model achieves very good accuracy. The mean error deviation between analytical and simulation results is at most 6% for LRU, 6% for the LFU, and 10% for the SIZE stack‐based algorithms. Copyright © 2009 John Wiley & Sons, Ltd.