Study of sub-word acoustical models for <Emphasis Type="Italic">Kannada</Emphasis> isolated word recognition system

The speech recognition system basically extracts the textual information present in the speech. In the present work, speaker independent isolated word recognition system for one of the south Indian language—Kannada has been developed. For European languages such as English, large amount of research has been carried out in the context of speech recognition. But, speech recognition in Indian languages such as Kannada reported significantly less amount of work and there are no standard speech corpus readily available. In the present study, speech database has been developed by recording the speech utterances of regional Kannada news corpus of different speakers. The speech recognition system has been implemented using the Hidden Markov Tool Kit. Two separate pronunciation dictionaries namely phone based and syllable based dictionaries are built in-order to design and evaluate the performances of phone-level and syllable-level sub-word acoustical models. Experiments have been carried out and results are analyzed by varying the number of Gaussian mixtures in each state of monophone Hidden Markov Model (HMM). Also, context dependent triphone HMM models have been built for the same Kannada speech corpus and the recognition accuracies are comparatively analyzed. Mel frequency cepstral coefficients along with their first and second derivative coefficients are used as feature vectors and are computed in acoustic front-end processing. The overall word recognition accuracy of 60.2 and 74.35 % respectively for monophone and triphone models have been obtained. The study shows a good improvement in the accuracy of isolated-word Kannada speech recognition system using triphone HMM models compared to that of monophone HMM models.     

Simulation and overall comparative evaluation of performance between different techniques for high band feature extraction based on artificial bandwidth extension of speech over proposed global system for mobile full rate narrow band coder

This paper addresses a novel approach to investigate, study and simulate computation of high band (HB) feature extraction based on linear predictive coding (LPC) and mel frequency cepstral coefficient (MFCC) techniques. Further, HB features are embedded into encoded bitstream of proposed global system for mobile (GSM) full rate (FR) 06.10 coder using joint source coding and data hiding before being transmitted to receiving terminal. At receiver, HB features are extracted to reproduce HB portion of speech and for the same different extension of excitation techniques are applied and their results evaluated in terms of quality (intelligibility and naturalness) and bandwidth. MATLAB based e-test bench is created for implementing the proposed artificial bandwidth extension (ABE) coder following series of simulations, that are carried out to discover and gain insight about the performance of it using subjective [mean opinion score (MOS)] and objective [perceptual evaluation of speech quality (PESQ)] analysis. The results obtained for both the analyses advocate that proposed ABE coder outperforms proposed GSM FR NB (legacy GSM FR) coder. While the fact remains that, compared to LPC based parameterizations over ABE coder, MFCC parameterization results in higher speech intelligibility which is evident from obtained slightly better PESQ and MOS scores.     

Which algorithm should I choose: An evolutionary algorithm portfolio approach

Many good evolutionary algorithms have been proposed in the past. However, frequently, the question arises that given a problem, one is at a loss of which algorithm to choose. In this paper, we propose a novel algorithm portfolio approach to address the above problem for single objective optimization. A portfolio of evolutionary algorithms is first formed. Covariance Matrix Adaptation Evolution Strategy (CMA-ES), History driven Evolutionary Algorithm (HdEA), Particle Swarm Optimization (PSO2011) and Self adaptive Differential Evolution (SaDE) are chosen as component algorithms. Each algorithm runs independently with no information exchange. At any point in time, the algorithm with the best predicted performance is run for one generation, after which the performance is predicted again. The best algorithm runs for the next generation, and the process goes on. In this way, algorithms switch automatically as a function of the computational budget. This novel algorithm is named Multiple Evolutionary Algorithm (MultiEA). The predictor we introduced has the nice property of being parameter-less, and algorithms switch automatically as a function of budget. The following contributions are made: (1) experimental results on 24 benchmark functions show that MultiEA outperforms (i) Multialgorithm Genetically Adaptive Method for Single Objective Optimization (AMALGAM-SO); (ii) Population-based Algorithm Portfolio (PAP); (iii) a multiple algorithm approach which chooses an algorithm randomly (RandEA); and (iv) a multiple algorithm approach which divides the computational budget evenly and execute all algorithms in parallel (ExhEA). This shows that it outperforms existing portfolio approaches and the predictor is functioning well. (2) Moreover, a neck to neck comparison of MultiEA with CMA-ES, HdEA, PSO2011, and SaDE is also made. Experimental results show that the performance of MultiEA is very competitive. In particular, MultiEA, being a portfolio algorithm, is sometimes even better than all its individual algorithms, and has more robust performance. (3) Furthermore, a positive synergic effect is discovered, namely, MultiEA can sometimes perform better than the sum of its individual EAs. This gives interesting insights into why an algorithm portfolio is a good approach. (4) It is found that MultiEA scales as well as the best algorithm in the portfolio. This suggests that MultiEA scales up nicely, which is a desirable algorithmic feature. (5) Finally, the performance of MultiEA is investigated on a real world problem. It is found that MultiEA can select the most suitable algorithm for the problem and is much better than choosing algorithms randomly.     

Modeling of microstrip antennas using neural networks techniques: A review

Artificial neural networks modeling have recently acquired enormous importance in microwave community especially in analyzing and synthesizing of microstrip antennas (MSAs) due to their generalization and adaptability features. A trained neural model estimates response very fast, which is nearly equal to its measured and/or simulated counterpart. Thus, it completely bypasses the repetitive use of conventional models as these models need rediscretization for every minor changes in the geometry, which itself is a time‐consuming exercise. The purpose of this article is to review this emerging area comprehensively for both analyzing and synthesizing of the MSAs. During reviewing process, some untouched cases are also observed, which are essentially required to be resolved for antenna designers. Unique and efficient neural networks‐based solutions are suggested for these cases. The proposed neural approaches are validated by fabricating and characterizing of the prototypes too. © 2015 Wiley Periodicals, Inc. Int J RF and Microwave CAE 25:747–757, 2015.     

Multiresolution approach for multiple human detection using moments and local binary patterns

Human detection is a central problem in development of any surveillance application. In this study, we present a simple and efficient, multi-resolution gray scale invariant approach for multiple human detection. The multiresolution is important for objects of different size and gray scale invariance is important due to uneven illumination and within-class variability. The proposed method is based on integration of central moments upon multi-resolution gray scale invariant local binary patterns operator. Since, the local binary patterns operator is invariant against different resolutions of space scale and monotonic change in gray scale, therefore the proposed method is robust in terms of variations in space scale as well as gray scale. Another advantage is high computational accuracy of the method due to use of moment operator which enhances the efficiency of the proposed method. Moreover, the proposed method is simple, as these operations can be performed within a few steps in a small neighborhood and a lookup table. The proposed method is tested on multiple human images and experimentally found appropriate for multiple human detection. The proposed method has been evaluated over two datasets, one is our own created dataset and the other is standard INRIA human detection dataset. Experimental results obtained from the proposed method demonstrate that better discrimination can be achieved for human and non-human objects in real scenes.     

Structure based virtual screening-driven identification of monastrol as a potent urease inhibitor

Virtual screening uses computer based methods to discover new ligands on the basis of biological structures. Among all virtual screening methods structure based docking has received considerable attention. In an attempt to identify new ligands as urease inhibitors, structure-based virtual screening (SBVS) of an in-house database of 10,000 organic compounds was carried out. The X-ray crystallographic structure of Bacillus pasteurii (BP) in complex with acetohydroxamic acid (PDB Code 4UBP) was used as a protein structure. As a starting point, ~10,000 compounds of our in-house database were analyzed to check redundancy and the compounds found repeated were removed from the database. Finally 6993 compounds were docked into the active site of BP urease using GOLD and MOE-Dock software. A remarkable feature of this study was the identification of monastrol, a well-known KSP inhibitor already in clinical trials, as a novel urease inhibitor. The hits identified were further evaluated by molecular docking and on examination of the affinity predictions, twenty-seven analogs of monastrol were synthesized by a multicomponent Biginelli reaction followed by their in vitro screening as urease inhibitors. Finally twelve compounds were identified as new urease inhibitors. The excellent in vitro activity suggested that these compounds may serve as viable lead compounds for the treatment of urease related problems.     

A cross-layer TCP enhancement in QoS-aware mobile ad hoc networks

With the increase of multimedia traffic over the Internet, current network protocols are largely concerned with the QoS requirements of delay-sensitive applications. In Mobile Ad-hoc Networks (MANETs), the majority of protocols developed to date provide QoS mechanisms by assigning high priority to delay-sensitive applications. While today’s Internet traffic is still dominated by TCP-based applications, the negative effects of the IEEE 802.11e service differentiation scheme on TCP performance in the presence of high priority traffic have not been adequately addressed in the literature. In this paper we first evaluate the performance of TCP in 802.11e MANETs when competing with high priority VoIP traffic. We then propose a novel TCP-friendly scheme, called IEDCA, to improve IEEE 802.11e EDCA mechanism. Our simulation-based performance study demonstrates that our proposed scheme IEDCA not only improves the performance of TCP significantly, it also facilitates the voice traffic transmission.     

Femtocell based green power consumption methods for mobile network

This paper presents the analytical models of power consumption in macrocell, microcell, picocell and femtocell based networks. Five case studies are presented in this paper where macrocells, microcells, picocells and femtocells are deployed based on the number of mobile subscribers present in a region, mobile user traffic in that region and the area of the region where cellular coverage has to be provided. A comparative study is performed between the power consumption by the base stations in each of these five cases and that of the only macrocell based network. The simulation results demonstrate that using each of these five strategies the power consumption by the base stations can be minimized than that of only macrocell based network. Based on the power consumption by the base stations in these five schemes, we have categorized the networks into five classes, A, B, C, D and E, each of which contains cells of different types to reduce power consumption to achieve green cellular network.