← Back to Archives
VOLUME 4, ISSUE 1, JANUARY 2015
Music emotion recognition using support vector machines and regression approach
T.N.CHARANYA, R.VIJAYALAKSHMI
To Access Remote Desktop using Windows Azure Cloud Computing
Saurabh Kapoor, Prof. Ashok Verma, Prof Ajay Lala
Synthesis and analysis of Fourier and format number transform operator on FPGA
Dr.Godbole B. B, Miss.PanditMadhuri D
Comparative Analysis of AM and MZ Modulator in 2Ă—100 Gbps Based WDM Optical Network
HimaniRathore, Ankita R. Mowar, Dr.SoniChanglani
Rule-Knowledge Based Algorithm for Event Extraction
Prashant G. Desai, Sarojadevi H., Niranjan N. Chiplunkar
Supporting privacy protection in personalized web search for knowledge mining
Mr. Stibu Stephen, Mr. A.Venugopal
Performance of Mobility Models with different Routing Protocols by using Simulation Tools for WSN: A Review
Megha Jain, V. K. Patle, Sanjay Kumar
Secure and Efficient Data Transmission for Cluster-Based Wireless Sensor Networks
Sanghamitra Panda, Satyanarayana Gandi, Amarendra Kothalanka
Prioritization of Functional Test Suites Using Closed Dependency Structures
C.VijayaKumar, M.S.Kokila, N.RajaSekaran
Qualified Breakdown of Wi-Fi with and without standard in Wireless Networks – Onus on Throughput
Bhat Geetalaxmi Jairam, D. V. Ashoka
Development of Filipino Phonetically-balanced Words and phoneme-level Hmms
Arnel C. Farjardo, Yoon-Joong Kim
Abstract
Music emotion recognition using support vector machines and regression approach
T.N.CHARANYA, R.VIJAYALAKSHMI
Abstract: Content-based retrieval has emerged in the face of content explosion as a promising approach to information access. In this paper, we focus on the challenging issue of recognizing the emotion content of music signals, or music emotion recognition (MER). Specifically, we formulate MER as a regression problem to predict the arousal and valence values (AV values) of each music sample directly. Associated with the AV values, each music sample becomes a point in the arousal-valence plane, so the users can efficiently retrieve the music sample by specifying a desired point in the emotion plane. Because no categorical taxonomy is used, the regression approach is free of the ambiguity inherent to conventional categorical approaches. To improve the performance, we apply principal component analysis to reduce the correlation between arousal and valence, and RReliefF to select important features. An extensive performance study is conducted to evaluate the accuracy of the regression approach for predicting AV values. The best performance evaluated in terms of the R2 statistics reaches 58.3% for arousal and 28.1% for valence by employing support vector machine as the regressor. We also apply the regression approach to detect the emotion variation within a music selection and find the prediction accuracy superior to existing works. A group-wise MER scheme is also developed to address the subjectivity issue of emotion perception.
Keywords: Music emotion recognition (MER), arousal, Valence, regression, support vector machine.
Keywords: Music emotion recognition (MER), arousal, Valence, regression, support vector machine.
Abstract
Feature selection for multiclass classification and regression techniques
N. Lakshmi, N. Deepika
Abstract: Online Multiclass Classification (OMC) performs the heterogeneous domain from complex data of completely diverse feature representation. OMC algorithm investigates the problem of heterogeneous domain and regression problems. Most existing studies of online learning divide the online feature selection into two parts. i) Learning with full input and ii) Learning with partial input. To address this limitation, we investigate the heterogeneous and regression problem in which an online learner is allowed to maintain a classifier with limited number of features. The key challenge of OMC algorithm is how to maintain the multiclass classification and regression using the active features. We attempt to tackle this challenge by studying on line feature selection and truncation techniques. We present OMC, Novel algorithm to solve the problems and give their performance analysis. We evaluate the performance of the proposed algorithms for on line learner using different domain and demonstrate their applications in real world problems including image classifications and analysis of bio informatics. Encouraging results validate the efficiency of our techniques.
Keywords: online multiclass classification, online Learning, Large scale data mining, Big data analytics.
Keywords: online multiclass classification, online Learning, Large scale data mining, Big data analytics.
Abstract
Security on cloud and pass
Mr Karthik.B, Mrs Devi selvam
Abstract: The term "cloud computing" has been mentioned for just under two years in relation to services or infrastructural resources, which can be contracted over a network. Cloud computing is a flexible, cost-effective, and proven delivery platform for providing business or consumer it services over the internet. However, cloud computing presents an added level of risk because essential services are often outsourced to a third party, which makes it harder to maintain data security and privacy, support data, service availability and demonstrate compliance. we analyze many technologies, it also inherits their security issues identifying the main vulnerabilities in this kind of systems and the most important threats found in the literature related to cloud computing and its environment as well as to identify and relate vulnerabilities and threats with possible solutions. Cloud computing is surrounded by many security issues like securing data, and examining the utilization of cloud by the cloud computing vendors. Cloud computing has brought lots of security challenges for the consumers and service providers. The study aims to identify the most vulnerable security threats in cloud computing, which will enable both end users and vendors to know about the key security threats associated with cloud computing. It will enable researchers and security professionals to know about users, vendors concerns and also to critical analysis about the different security models and tools proposed.
Keywords: Cloud security, Iaas, Paas, cloud computing
Keywords: Cloud security, Iaas, Paas, cloud computing
Abstract
DWT-OFDM Diversity for TSV-Model Based 60 GHz WPAN System
C N Deshmukh, V T Ingole
Abstract: In the context of WLAN (Wireless Local Area Network) and WPAN (Wireless Personal Area Network) systems for High Data Rate (HDR) wireless communications, the unlicensed frequency band available in the millimeter wave region has become more and more attractive. Currently focus is on use of OFDM system to cater for increased data rate of wireless medium with good performance. Diversity techniques play an important role in achieving higher performance level for limited power wireless systems. Wavelet analysis has some strong advantages over Fourier analysis, as it allows a time-frequency domain operation, allowing optimal resolution and flexibility. Wavelets have been satisfactorily used in almost all the fields of wireless communication systems including OFDM which is a strong candidate for next generation of wireless system. This paper proposes a DWT-OFDM Diversity to achieve better performance in terms of SNR and bit error rate (BER) for TSV model based channel at 60 GHz. The performances of different discrete wavelets for channels defined by IEEE 802.15.4a are analyzed. The results indicate better BER performance in case of lower order wavelet.
Keywords: DWT, OFDM, TSV Model, WPAN, WLAN, FFT, SNR, BER, ISI
Keywords: DWT, OFDM, TSV Model, WPAN, WLAN, FFT, SNR, BER, ISI
Abstract
To Access Remote Desktop using Windows Azure Cloud Computing
Saurabh Kapoor, Prof. Ashok Verma, Prof Ajay Lala
Abstract: This paper describe about an era of cloud computing platform in terms of Azure cloud platform and services. When choosing remote desktop control software in business environment following important decision to be made possible in cloud computing such as In house hosting, self hosting remote desktop for Software as a service tools. It defines the WCF RESTful services and Azure Appfabric Service Bus working together to allow thin client to access remote machine using a browser. It may consist of Azure service bus to communicate with application to share a target machine screen. The application running on a machine which helps us to allow the user to control and view the target machine desktop screen from browser.
Keywords: Cloud computing, WCF Service, REST Service,Azure Appfabric service bus.
Keywords: Cloud computing, WCF Service, REST Service,Azure Appfabric service bus.
Abstract
FPGA implementation of cordic algorithm used in DDS based modulators
Godbole B. B., Nikam R.H.
Abstract: The modern communication systems and software radio based applications demands smart transceivers, consisting of only an antenna and a fully programmable circuit with digital modulators and demodulators. A basic communication system’s transmitter modulates the amplitude, phase or frequency proportional to the signal being transmitted. An efficient solution (that doesn’t require large tables/memory) for realizing universal modulator is CORDIC (CO-ordinate Rotation Digital Computer) algorithm. The CORDIC algorithm is used in the rotation mode, to convert the coordinates from polar mode to rectangular mode. CORDIC is a versatile algorithm widely used for VLSI implementation of digital signal processing applications. This paper presents how to use CORDIC to implement different communication subsystems that can be found in software defined radio. Specifically, it shows how to use CORDIC to implement direct digital synthesizers and ASK, PSK, FSK Modulators. The focus of this paper is to analysis and simulation of ASK, FSK, PSK modulation scheme using Direct Digital Synthesizer having CORDIC algorithm at the place of ROM Look up Table. CORDIC Algorithm provides many significant advantages over Conventional ROM Look up Table.
Keywords: Software Defined Radio, CORDIC algorithm, DDS, ASK, FSK, PSK.
Keywords: Software Defined Radio, CORDIC algorithm, DDS, ASK, FSK, PSK.
Abstract
Synthesis and analysis of Fourier and format number transform operator on FPGA
Dr.Godbole B. B, Miss.PanditMadhuri D
Abstract: Reconfiguration is an essential part of Software Radio (SR) technology. In this SR context, the Fast Fourier Transform (FFT) operator is defined as a common operator for many classical telecommunications operations. The systems are designed with a new architecture for this operator that makes it a device intended to perform two different transforms. The first one is the Fast Fourier Transform (FFT) used for the operations in OFDM based communication. The second one is the Fermat Number Transform (FNT) used for the finite operations in the Galois Field (GF). This operator can be reconfigured to switch from an operator dedicated to compute the FFT to the FNT in the Galois Field.
Keywords: Software Rodio, FPGA, Reconfigurable operator, FFT,FNT.
Keywords: Software Rodio, FPGA, Reconfigurable operator, FFT,FNT.
Abstract
Comparative Analysis of AM and MZ Modulator in 2Ă—100 Gbps Based WDM Optical Network
HimaniRathore, Ankita R. Mowar, Dr.SoniChanglani
Abstract: Multiplexing is widely employed due to its capability to increase transmission capacity and to reduce system costs. The system features multiplexing of the basebands in electrical domain as well as multiplexing in optical domain (WDM). The design is proposed for 10- users, each one assigned a different RZ duty cycle and with a data rate of 20 Gbps. In this work, optical wavelength Division Multiplexing is done on wavelengths 1550nm and 1552nm. There are two modulation schemes AM and MZ are compared for 2Ă—100 Gbps Based WDM Optical Network. The AM shows better performance as compared to MZ modulation scheme, AM transmitted successfully upto 78km where as MZ up to 75km.
Keywords: WDM, BER, AM, Mach-Zehnder
Keywords: WDM, BER, AM, Mach-Zehnder
Abstract
Rule-Knowledge Based Algorithm for Event Extraction
Prashant G. Desai, Sarojadevi H., Niranjan N. Chiplunkar
Abstract: The amount of electronic data being produced and communicated online is rapidly increasing. Most of these electronic documents contain lot of important & interesting information in unstructured format. Various approaches are used in the literature to automate the process of event extraction so as to conserve time and effort. This paper proposes an algorithm to automate the task of information extraction. We demonstrate that machine learning can be used for extracting event from the unstructured text such as dates places and subject of interest.
Keywords: Automation, Machine Learning, Unstructured Text, Information Extraction, Natural Language.
Keywords: Automation, Machine Learning, Unstructured Text, Information Extraction, Natural Language.
Abstract
Supporting privacy protection in personalized web search for knowledge mining
Mr. Stibu Stephen, Mr. A.Venugopal
Abstract: With increasing number of websites the Web users are increased with the massive amount of data available in the internet which is provided by the Web Search Engine (WSE). The aim of the WSE is to provide the relevant search result to the user with the behavior of the user click were they performed. WSE provide the relevant result on behalf of the user frequent click based method. From this method no assurance to the user privacy and also no securities were providing to their data. Hence users were afraid for their private information during search has become a major barrier. They were many techniques were proposed by researchers most of that based on the server side, it has provide less security. For minimizing the privacy risk here we propose the client side based technique with the combination of Greedy method to prevent the user data that we applied in Knowledge mining area.
Keywords: Web Search Engine, personalized search, user query logs, content search and privacy preserving.
Keywords: Web Search Engine, personalized search, user query logs, content search and privacy preserving.
Abstract
Estimating and determining the position of target in WSN’s
K.Harshini, Deepika Vodnala
Abstract: In the proposed paper the major issue is the target tracking in wireless sensor networks and to obtain exact position of the target as to be measured so that we can expect improvement in the tracking resolution. so there is need of finding accurately sensor and targets trajectory path as to be studied with special care and captured at regular basis. We also studied correlation properties and sensitivity in mobile sensors from the system parameters and maintained good resolution for tracking of mobile sensors at different speeds. Our simulation results gave satisfactory system parameters like sensor density, sensor and target mobility compared static sensor network environment.
Keywords: Component, Mobile sensor networks, sensor range, resolution, Mobility Model
Keywords: Component, Mobile sensor networks, sensor range, resolution, Mobility Model
Abstract
A Review of Computerized Payroll System
Kritika Mahajan, Shilpa Shukla, Nitasha Soni
Abstract: Payroll is a critical operation for every organization to pay employee accurately their salary and enrollments on time.[1]The idea of taking control of employees pay calculations are quite tedious if done manually and require more effort and time mainly for big organizations. Hence if this process is automated, it would be of great benefit as it would require less time to calculate the salary of the employees. The software for payroll management system service on the cloud is provided as a solution in this paper.[2] This system provides multiple user data access. Each user like employee or HR or admin can login into the software by writing username and password which are allocated to them from the company. It involves keeping track of hours worked and is capable of keeping a record of employee data including their pay, allowances, deductions and taxes on monthly bases so that fresh definitions are reflected from the month onwards, which leaves all the past data intact.[3] The proposed payroll system is advantageous as it provides a user friendly environment and also increases security and minimizes human calculation errors.
Keywords: Cloud Computing, Payroll System, 3-tier architecture, Payroll Working Process, Manual Payroll, Computerized Payroll, JavaScript, CSS, HTML, Ajax and JSON.
Keywords: Cloud Computing, Payroll System, 3-tier architecture, Payroll Working Process, Manual Payroll, Computerized Payroll, JavaScript, CSS, HTML, Ajax and JSON.
Abstract
A QoS-oriented distributed routing protocol for hybrid wireless networks
Mathumathi, Manjusha.T.M
Abstract: With the increasing level of wireless communication in today’s environment, people often required QOS for sharing their data between the nodes. For affording QOS to the user, many researchers proposed a very few methods to provide QoS guaranteed routing for hybrid networks, they strive to improve the network capacity and reliability but they evade constrain in QOS. For this problem our main objective of this paper is to improve the QOS and efficiency of routing approach with constrains over hybrid wireless data streaming using QOS_DARP protocol and MAR. This aims to develop the QOS based reliable architecture against the hybrid wireless routing issues. The system also aims at providing both proactive and reactive solutions for effective routing. The goal of this paper is to providing efficient dynamic routing management to deal the challenges of data transmission and data streaming in hybrid mobile environments.
Keywords: Hybrid Wireless Network, QoS, QOS_DARP and MAP protocol.
Keywords: Hybrid Wireless Network, QoS, QOS_DARP and MAP protocol.
Abstract
Performance of Mobility Models with different Routing Protocols by using Simulation Tools for WSN: A Review
Megha Jain, V. K. Patle, Sanjay Kumar
Abstract: Research on the mobility models for wireless sensor networks has been an interesting area of networking. As a result, researchers over the past few years have measured the performance of mobility models by applying protocols through various simulation tools for wireless sensor networks. Wireless sensor network is a group of sensing devices which are geographically distributed that monitor the condition around it such as physical or environmental which communicates wirelessly to cooperatively pass data to the main location. The goal of the proposed work is to provide researchers with clear idea about the available mobility models for wireless sensor network performance which can be investigated on various available simulation tools with several metrics of sensor networks. Hence, we attempt to provide an overview of the current research status for wireless sensor network mobility models.
Keywords: Mobility Model, WSN, Performance Metrics, Simulation Tools, Routing Protocols.
Keywords: Mobility Model, WSN, Performance Metrics, Simulation Tools, Routing Protocols.
Abstract
The Indian kaleidoscope: emerging trends in M-Commerce
S.Muthukumar, Dr.N.Muthu
Abstract: India is the second largest cellular market in the world after China, with a massive subscriber base of 867.80 million, as of March 2013. Majority of smartphone users are still on 2G network. Budget 4G smartphones coupled with affordable plans, can very well drive 4G growth in India. The most obvious mobile commerce trend is further development. Yearly m-commerce sales are forecasted to increase fourfold billion in the next few years. Businesses are beginning to realize that m-commerce is key to enhance their brand, boost sales, and keep up with competitors.India’s retail market is expected to cross 1.3 trillion USD by 2020 from the current market size of 500 billion USD. Modern retail with a penetration of only 5% is expected to grow about six times from the current 27 billion USD to 220 billion USD, across all categories and segments.India is set to witness proliferation of the fourth-generation wireless data services, or 4G services shortly with slashed data plans. Being the second largest mobile market in the world, India needs to take its place in the forefront of providing innovative services and applications to its citizens. Recent eMarketer study, by the year 2017 more than 25% of all online retail transactions will happen in the mobile paradigm. Adweek explains that statistic with information that 18-34 year olds are very likely to use their mobile devices as a shopping tool. Their process is to visit their favorite retail stores not to shop but to view a product and compare prices, and then to compare prices at various online locations using their phones. They then buy the product using their mobile device. The future looks very bright for mobile commerce, although businesses are still experimenting with how to use the mobile commerce concept to their best advantage.
Keywords: m-commerce, trends, future, e –tail
Keywords: m-commerce, trends, future, e –tail
Abstract
DSP in image processing
Pratibha Sharma
Abstract: Vision extracts useful information from images. Reconstructing the three-dimensional structure of our environment and recognizing the objects that populate it are among the most important functions of our visual system. Computer vision researchers study the computational principles of vision and aim at designing algorithms that reproduce these functions. Vision is difficult: the same scene may give rise to very different images depending on illumination and viewpoint. Typically, an astronomical number of hypotheses exist that in principle have to be analyzed to infer a correct scene description. Moreover, image information might be extracted at different levels of spatial and logical resolution dependent on the image processing task. Knowledge of the world allows the visual system to limit the amount of ambiguity and to greatly simplify visual computations. We discuss how simple properties of the world are captured by the Gestalt rules of grouping, how the visual system may learn and organize models of objects for recognition, and how one may control the complexity of the description that the visual system computes.
Keyword: Image Processing, Spatial and Logical Resolution, Visual system.
Keyword: Image Processing, Spatial and Logical Resolution, Visual system.
Abstract
Optimal Game Theory for Network Security using IPDRS Engine
Shyam Chandran P, Resmi .A.M
Abstract: With the tremendous growth of network technology, network attackers are also rationally increased to disrupt the activities and hacking the data from the network users by the intruder. To provide the security from the intruder is most important one. Many researchers were proposed a different approach for providing security they does not tackle the problem, it leads an untruthful in the network. In existing system they propose a game based intrusion request and response process by game play of header node to access the data by their correct play, from this if the real header play it in wrong method means all the client user under the header also suffer from the access of the data. To tackle this issue we enhanced the game theory based on the individual play in the network to avoid the suffering of misplayed game played by the header using the IPDRS Engine.
Keywords: Network Security, Intrusion Response System, Game Theory, Optimal Game and IPDRS Engine.
Keywords: Network Security, Intrusion Response System, Game Theory, Optimal Game and IPDRS Engine.
Abstract
Online content optimization with effective event monitoring approach
Sajithra.N, Priyadarshini.D
Abstract: Content optimization has widely used in personalized search engines for better personalized results. It can be generated by the user actions and events on the web search engine. User interaction on the page plays a vital role in recommender systems. Previous studies on recommender systems mainly focused on modeling techniques and feature development, this content optimization is based on general behavior analysis algorithm. It provides user action analysis is critical for a recommender system. The system proposes a novel implicit user feedback and event monitoring schemes for efficient content optimization. For this our system proposes PCO approach. But user interactions in real- world Web applications are unlikely to be as ideal as those assumed by previously proposed models. Our proposed system builds an online dynamic learning framework for personalized recommendation. The main contribution in this paper is an approach of personalizing users' searches to achieve better search result which is based on event monitoring and personalized content optimized search.
Keywords: Content optimization, web search engine, Event monitoring, PCO and DKMC.
Keywords: Content optimization, web search engine, Event monitoring, PCO and DKMC.
Abstract
Secure and Efficient Data Transmission for Cluster-Based Wireless Sensor Networks
Sanghamitra Panda, Satyanarayana Gandi, Amarendra Kothalanka
Abstract: Secure and efficient data transmission is a critical issue for cluster-based wireless Sensor Networks (WSNs). In Cluster–based WSNs authentication of users is a very Important issue .So, by authenticating the sent user and the destination user , we can achieve the security and efficiency of data over CWSNs. To provide security of data and authentication of user we proposed a technique where we are implementing two concepts for performing those operations. The first one is identity based signature (IBS) for verification of user generated by the verifier and second one is a key is xor operated with the data and get the cipher and then binary level technique for encryption and decryption of the original message. The binary level technique converts the plain text into binary form and then splits the data into blocks and assign values to it based on identification mark (IM) technique which depends upon the length of the binary digits, then these are divided into two level, 1st level is 2 bit and 2nd level is 4 bit . Then at the receiver user the Cipher text will be decrypted by using the reverse technique and the destination user will get the original message. By providing those techniques we can improve efficiency, security overhead and energy consumption.
Keywords: Identity based Signature (IBS), shared key generation, User authentication, message encryption and decryption.
Keywords: Identity based Signature (IBS), shared key generation, User authentication, message encryption and decryption.
Abstract
Prioritization of Functional Test Suites Using Closed Dependency Structures
C.VijayaKumar, M.S.Kokila, N.RajaSekaran
Abstract: Test cases organize the whole testing process. If the test cases are prepared with the requirements of a particular system then it helps in testing whether the requirements are fulfilled or not. A defect is an error in coding or logic that causes a program to malfunction or to produce incorrect/unexpected results. Increasing the rate of fault detection can provide earlier feedback to system developers, improving fault fixing activity and ultimately software delivery. The system uses the knowledge based and model based prioritization to prioritize the test case. So the efficiency of the test case is increased and the running time for the test cases is decreased. When using coarse grained technique the fault is identified easily. Due to the earlier feedback to system developers which makes the software delivery earlier.
Keywords: Test Case, Prioritization, Error, Dependency Structures.
Keywords: Test Case, Prioritization, Error, Dependency Structures.
Abstract
Qualified Breakdown of Wi-Fi with and without standard in Wireless Networks – Onus on Throughput
Bhat Geetalaxmi Jairam, D. V. Ashoka
Abstract: Wireless networking equipments available supports varying levels of industry communication standards. At present the IEEE 802.11 b/g standards are widely accepted throughout the industry and provide the necessary balance of range, network throughput and support for device mobility to effectively serve most needs of university community. The authors have concentrated on through put of different networks consisting of 4,5,6 and 7 nodes. Authors are trying to justify that applying standards to wireless network will improve its throughput than without applying any standard. Authors also made comparative analysis of throughput obtained for topology consisting of different number of nodes with standard and without standard and ended up with the conclusion that throughput obtained with standard 802.11g is more than 802.11b.
Keywords: Audio Frequency, Radio Frequency, Wireless fidelity, unregulated signal frequency.
Keywords: Audio Frequency, Radio Frequency, Wireless fidelity, unregulated signal frequency.
Abstract
Face Recognition System Using PCA and DCT in HMM
SamerKais Jameel
Abstract: The speed of procedures became necessities during recent years; therefore using computers turn out to be the most important factors to increase the speed of implementation especially in security aspect such as recognition of people.There are a lot of waysto recognize the people face recognition is one of them. In this work the details of the face have been taken as blocks and Discrete Cosine Transform (DCT) is used, applied on face image’s blocks. Then without doing inverse DCT Principal Component Analysis (PCA) is applied directly for dimensionality reduction this makes the system very fast. Olivetti Research Laboratory (ORL) database of faces had been used to obtain the results.Each face is considered as a numerical sequence (blocks) that can be easily modelled by HMM. On 400 face images of the (ORL) face database the system has been examined. The experiments showed a recognition rate of 95.211%, using half of the images for training.
Keywords: Face Recognition, Hidden Markov Model, Discrete Cosine Transform (DCT), Principal Component Analysis (PCA).
Keywords: Face Recognition, Hidden Markov Model, Discrete Cosine Transform (DCT), Principal Component Analysis (PCA).
Abstract
A Novel Optimum Real-Time Color Reduction on FPGA Based on Swarm Intelligence
Khairy Assar
Abstract: Data clustering is a popular approach for automatically finding set of objects into a specific number of clusters. Clustering is largely used in many fields including text mining, information retrieval and pattern grouping. Particle Swarm Optimization (PSO) is a population-based optimization algorithm modelled after the simulation of social behaviour of bird flocks and widely used for optimize problem solving. In clustering problem PSO gives optimal solution but takes long time (so called iterations) to find the optimum solution. The hybrid PSO and K-means algorithm is developed to automatically detect the cluster centers of geometrical structure data sets. The proposed algorithm gives the benefits for each of two-merged algorithms. K-means is fast algorithm, PSO optimize the solution. The implementation of the hybrid K-means PSO structure is realized in hardware. The clustering based on hybrid K-means PSO architecture is described by different technique for hardware description (i.e. block diagram) and implemented on field programmable gate array (FPGA). Its feasibility is verified by experiments. Results show that the proposed architecture implemented on the FPGA has a good clustering technique especially for testing with color reduction for true color video.
Keywords: Clustering, K-means, Color image reduction, Particle Swarm Optimization (PSO), field programmable gate array (FPGA), and Real Time Video Color Reduction.
Keywords: Clustering, K-means, Color image reduction, Particle Swarm Optimization (PSO), field programmable gate array (FPGA), and Real Time Video Color Reduction.
Abstract
Development of Filipino Phonetically-balanced Words and phoneme-level Hmms
Arnel C. Farjardo, Yoon-Joong Kim
Abstract: In this paper, we developed two sets of phonetically balanced word (PBW) lists in Filipino and two sets of phoneme-level HMMs (Hidden Markov Model). Two PBW lists were based on textbooks used in public school in Philippines and used to develop speech corpus with the fifty speakers of 25 males and 25 females. In a 2-syllable word list (PBW2), an average accuracy rate of 88.95% for speaker dependent and 82.57% for speaker independent test were achieved. For 3-syllable word list (PBW3), the recognizer achieved an accuracy rate of 90.28% for speaker dependent and 83.30% for speaker independent test.
Keywords: Filipino phonetically balanced words, Filipino word corpus, Hidden Markov Model, Automatic Speech Recognition .
Keywords: Filipino phonetically balanced words, Filipino word corpus, Hidden Markov Model, Automatic Speech Recognition .