Kākināda, India
Kākināda, India

Time filter

Source Type

Suneetha Rani T.,JNTUK | Krishna Prasad M.H.M.,JNTUK
International Journal of Systems Assurance Engineering and Management | Year: 2017

Identifying the cues for speech segments of speech data is an indispensable task in speaker clustering. The existing techniques perform the task of speech clustering without any prior knowledge of cluster tendency. Many techniques are investigated for finding a prior cluster tendency (CT). During the investigation, the visual access tendency (VAT) is recognized as a reasonable choice to find a cluster tendency. The speech clustering poses three important problems, which are as follows: modelling the speech data, cluster tendency, and effective speech clustering. Modelling is required for defining the shape of the speech segment based on the characteristics of speaker’s voice; hence it is useful for speech recognition. The GMM is a good choice for obtaining the precise model of speech data. Determining the number of speakers (or number of clusters) for the speech is known as cluster tendency. The quality of speech clustering depends on modelling and a prior clustering tendency. The classical algorithms [such as k-means, and minimum spanning tree (MST)-based-clustering] are merged with VAT for determining the effective clustering results along with a prior cluster tendency. We use linear subspace learning for representing the speech segments (or speech utterances) in a projected space of high-dimensional data. Various linear subspace learning techniques are used for improving the speech clustering results. The proposed approaches are hybrid approaches (i.e., k-means-CT, and MST–CT-based clustering), they use expensive steps. For this key reason, we propose another method, direct visualized clustering method, in which we derive the explicit speaker clustering results directly from VAT instead of using either k-means or MST-based clustering. We experimented the proposed methods on TSP speech datasets and done the comparative study for demonstrating the effectiveness of our work. © 2015, The Society for Reliability Engineering, Quality and Operations Management (SREQOM), India and The Division of Operation and Maintenance, Lulea University of Technology, Sweden.


Prasad K.S.,JNTUK
Lecture Notes in Electrical Engineering | Year: 2017

The limitation with the existing testing techniques is, if the test does not consider all the aspects of SRAM parameters, including parasitic memory effect, then it will result as an incomplete test. This paper presents a new parasitic extraction testing method for embedded SRAMs, employing defect-induced layout. The defect injection in a circuit is due to an open/short between wires, or missing contacts etc. In this work, only node-to-node short defects are considered. Our test results proved that using parasitic extraction method existing faults as well as undefined faults could be detected. The existing faults identified are Stuck At Fault, (SAF), Undefined Read Fault (URF), Read Destructive Fault (RDF), Undefined Write Fault (UWF), Random Read Fault (RRF), Incorrect Read Fault (IRF), and No Access Fault (NAF). The undefined faults identified are Bit-Line Delay Fault (BDF), Initialization Order Fault (IOF), Un Stabilized Write Fault (USWF), Un Stabilized Read Fault (USRF), and Write Before Access Fault (WBAF). In addition, the complete fault model dictionary is also an outcome of this work. © Springer Science+Business Media Singapore 2017.


Srinivasa Rao R.,JNTUK | Srinivasa Rao V.,JNTUK
International Journal of Electrical Power and Energy Systems | Year: 2015

This paper presents a generalized approach for determination of optimal locations for placement of Flexible AC Transmission Systems (FACTs) devices in the power system with an objective of reducing real power loss and to reduce overloading of the lines. An objective function involving above objectives is formulated and a detailed mathematical model for each objective is presented in terms of system parameters. Three FACT devices, namely, Unified Power Flow Controller (UPFC), Interline Power Flow Controller (IPFC), and Optimal Unified Power Flow Controller (OUPFC) which are capable of controlling both active and reactive power are considered in simulation and analysis of the networks. The parameters to be optimized have been identified and incorporated in the objective function for each device. Sensitivity analysis is used to locate optimal buses to place the FACTs devices in the network. Effectiveness of the approach is demonstrated on a 5 Bus and an IEEE 14 Bus systems for each FACT device. Simulation results obtained for each device using proposed approach are compared with those obtained in the literature. © 2015 Elsevier Ltd. All rights reserved.


Challa S.R.,JNTUK | Rama Koti Reddy D.V.,Andhra University
Advances in Intelligent Systems and Computing | Year: 2016

Image enhancement technique plays a vital role in image analysis. White-band images (WBI) and narrow band images (NBI) are the two types of images that are widely used in capsule endoscopy. It is a general practice to compress the images and process them at the chip level embedded in the capsule in order to minimize the power consumption. In the due course, it is often required to enhance the features of the image so that, most favorable information from the images can be extracted. Loss of naturalness in the image while processing for image enhancement is observed. In this paper, a technique called brightness enhancement followed by color restoration is suggested for images with non-linear illumination. The technique employs a bright pass filter which acts on reflectance and luminance forms of image obtained after decomposing of the original image. The simulation experiment is carried on both white-band images and narrow-band images. © Springer India 2016.


Narasaraju G.,JNTUK | Raju D.L.,Andhra University
Materials Today: Proceedings | Year: 2015

The application spectrum of low cost material reinforced metal matrix composites is growing rapidly in various engineering fields due to their superior mechanical properties. In the present study an attempt is made to explore the possibilities of reinforcing aluminium alloy (AlSi10Mg) with locally available inexpensive rice husk and fly ash for developing a new composite material. Hybrid Rice husk and Fly ash particles are added in Aluminium Alloy Matrix at 20% by weight with different proportion using Stir casting method. The fabricated cast specimens are characterized with Optical and Scanning Electron Microscopes. SEM study validates the presence of Rice husk and Fly ash in Aluminium matrix. The mechanical properties such as tensile strength, percentage elongation, and hardness are studied for all stir-cast specimens. A significant improvement is observed for 10% Rice husk+10% Fly ash reinforced Aluminium Alloy composite compared to as-cast specimens without reinforcement additions. © 2015 Elsevier Ltd.


Venkata Rao K.,Vignan University | Murthy B.S.N.,Gandhi Institute of Technology and Management | Mohan Rao N.,JNTUK
Measurement: Journal of the International Measurement Confederation | Year: 2013

The vibration is one of the intensive problems in boring process. Machining and tool wear are affected more by vibration of tool due to length of boring bar. The present work is to estimate the effect of cutting parameters on work piece vibration, roughness on machined surface and volume of metal removed in boring of steel (AISI1040). A laser Doppler vibrometer (LDV) was used for online data acquisition and a high-speed FFT analyzer used to process the AOE signals for work piece vibration. A design of experiments was prepared with eight experiments with two levels of cutting parameters such as spindle rotational speed, feed rate and tool nose radius. Taguchi method has been used to optimize the cutting parameters and a multiple regression analysis is done to obtain the empirical relation of Tool life with roughness of machined surface, volume of metal removed and amplitude of work piece vibrations. © 2013 Elsevier Ltd. All rights reserved.


Pavanipadmapriya B.,JNTUK
International Journal of ChemTech Research | Year: 2013

A simple, precise and sensitive procedure for simultaneous estimation of ciprofloxacin and tinidazole in two component tablets has been developed using 50% v/v glacial acetic acid. The absorbance maxima of ciprofloxacin and tinidazole are 281.7 nm and 310.9 nm respectively. Both the drugs obey Beer's Law in the concentration ranges employed for the proposed method. The results of analysis have been validated statistically and by the recovery studies. The method showed no interference by the compounds from each other when applied to formulation by means of UV-VIS spectroscopy and can be applied for routine simultaneous estimation of both drugs.


Gopi Krishna Rao P.V.,JNTUK | Subramanyam M.V.,Shantiram Engineering College | Satyaprasad K.,JNTUK
Advances in Intelligent Systems and Computing | Year: 2015

Proportional integral (PI) and proportional integral derivative (PID) controllers have been at the heart of control engineering practice for several decades. 95 % of the controllers employed in the industry are PI/PID. In process control, one often encounters systems described by transfer functions with time delays, which become transcendental functions. The design of the controller demands the rational transfer function approximation of the time-delay term. This paper focuses on the effect of time-delay approximation techniques, viz. Taylor series expansion and Padé approximation, on the structure and performance of PI/PID controllers designed with Internal Model Control (IMC). The performance of the PI/PID controllers was tested in simulation environment on various processes with time delay. For uniform comparison, the controllers were tuned to have a same robustness measure, in terms of maximum sensitivity (M S). The results indicate, irrespective of time-delay approximation considered, the controllers provide good set point tracking and poor disturbance rejection. © Springer India 2015.


Venkata Rao K.,Vignan University | Murthy B.S.N.,Gandhi Institute of Technology and Management | Mohan Rao N.,JNTUK
Measurement: Journal of the International Measurement Confederation | Year: 2014

Machining of stainless steel is difficult due to their hardening tendency. In boring of stainless steels, tool wear and surface roughness are affected by vibration of boring bar. In this paper, tool wear, surface roughness and vibration of work piece were studied in boring of AISI 316 steel with cemented carbide tool inserts. A Laser Doppler Vibrometer was used for online data acquisition of work piece vibration and a high-speed Fast Fourier Transform analyzer was used to process the acousto optic emission signals for the work piece vibration. Experimental data was collected and imported to artificial neural network techniques. A multilayer perceptron model was used with back-propagation algorithm using the input parameters of nose radius, cutting speed, feed and volume of material removed. The artificial neural network was used to predict surface roughness, tool wear and amplitude of work piece vibration. The predicted values were compared with the collected experimental data and percentage error was computed. © 2014 Elsevier Ltd. All rights reserved.


Srinivas R.,SSAIST | Ramana K.V.,JNTUK
Procedia Computer Science | Year: 2015

Spine is a conspicuous part of human body. It encounters a process of changes with age, stress and strain which originates various spinal problems like disc degeneration, disc herniation, and inter vertebral compression. Labeling plays a pivotal role in diagnosing spinal problems. This paper presents a unique method VertebraeStatistics Description Algorithm (VESTAL) to label cervical vertebrae and inter vertebral discs (IVDs). VESTAL uses statistical feature of cervical vertebrae and IVDs. A template is created by VESTAL by determining both posterior, anterior width, heights of cervical vertebrae and IVDs. The template which is formulated will be used for detecting vertebrae and IVDs. This VESTAL application is applied on 45 patients which brought out 225 MR images with 96% accurate results in labeling. © 2015 The Authors. Published by Elsevier B.V.

Loading JNTUK collaborators
Loading JNTUK collaborators