Call for Papers‎ > ‎

Volume 8, Issue 2, August 2011

Syllabification Algorithm based on Elicitation Method and Syllable Matching Rules for Malay Language [Full Text]

Rabiah A.Kadir, Hafiz Musa, Azreen Azman and M.Taufik Abdullah

In this paper, we present a new syllabification algorithm for Malay language. Syllabification is the process to extract
or  divide  syllable  from  words.  Syllabification  process  is  language  dependent  where  each  language  can  have  its  own  set  of syllable structure. Syllabication is an important component in speech synthesizer, speech recognition and transliteration system. Syllabification  algorithms  have  been  proposed  in  many  languages  including  English,  Spanish,  Myanmar,  Singhala,  Chinese and ect.  Unfortunately,  there  are  not  many information regarding  evaluation  of syllabification scheme for  Malay.  In  this paper, we propose an efficient algorithm based on elicitation method and syllable  rules matching.  In order to evaluate the algorithm, a prototype  has  been  developed  to  measure  the  accuracy  of  syllabification.  We  evaluate  our  method  using  Bernama,  Kamus Dewan and Overlap data collection. The syllable rules matching achieved 60.7% accuracy on BERNAMA collection, 77.4% on Kamus Dewan Collection and 71.6% on
Overlap collection.


Feature Reduction Techniques  For Image Steganalysis
[Full Text] 

J.Charlang and H.Sajedi

One of the factors that affects the success of steganalysis is the number and quality of the features that form the feature vectors. If  feature  is  irrelative  or  redundant,  or  the  feature  is  noisy  and  unreliable,  then  steganography  discovery  during  training  is  more  difficult. Attribute subset selection is the process of identifying and eliminating as much of the irrelative and redundant information as possible. Learning algorithms differ in the amount of emphasis they place on attribute selection.The selection of a little number of highly predictive features in order  to  eschew  over fitting  the  training  data.  Irrespective  of  how  a  learner  attempts  to  select  attributes  itself  or  ignores  the  topic,  attribute selection prior  to learning  can be useful. Reducing  the dimensionality  of  the feature reduces  the  size of the  supposition space and allows algorithms to act faster and more efficiently. In some cases, accuracy on future classification can be improved; in others, the result is a more compact, easily interpreted representation of the target concept. This paper presents a benchmark comparison of several attribute selection Techniques for reduction of feature vector dimensionality in blind image steganalysis. All the methods produce an attribute ranking, a useful devise for isolating the individual merit of an attribute. Attribute selection is achieved by cross-validating the attribute rankings with respect to a classification learner to find the best attributes. 


Design and Implementation of Digital Filters Using Bit-Serial Arithmetic & Bit-Parallel Arithmetic
[Full Text]

Shib Sankar Bhowmick, Debotosh Bhattacharjee, Mita Nasipuri and Dipak kumar Basu

Frequency  selective  digital  filters  are  widely  used  in  most  of  the  DSP  (digital  signal  processing)  applications.
Moreover these filters are dealing with high data rates, must also have a high throughput. In order to develop such systems, we need  efficient,  fast,  and  reliable  design  methods  that  support  low  cost  circuit  implementations  also.  To  design  such  circuits having  high  throughput  with  low  power  consumption,  some  suitable  algorithm  along  with  the  associated  circuitry  required  for digital  signal  processing  need  to  be  optimized  before  implementing  it  into  the  FPGA  kit.  In  this  present  work  design  and implementation  of  frequency  selective  digital  filters  using  both  bit  serial  [1]  and  bit  parallel  arithmetic  [1]  have  been  done.  In addition to that the  work also proposes some arithmetic transformations  in  design  of  digital  filters  which  further increases the throughput and reduces the power consumption. For implementing the desired filters, VHDL and FPGA have been chosen as a working  platform.  Here  three  different  digital  FIR  filters  have  been  realized  by  using  bit  serial  and  bit  parallel  arithmetic  and finally implemented by VHDL. These three types of filters are (i) Direct form, (ii) Cascade form and (iii) Lattice form.


High capacity audio watermarking based on wavelet transform
[Full Text]

D.Zanganeh and H.Sajedi

This paper suggests a novel high capacity robust audio watermarking algorithm by using the high frequency band of the wavelet decomposition, for which the human auditory system (HAS) is not very sensitive to alteration. The main idea is embedding two secret bits into a singlewavelet coefficient. The experimental results show that the method has very high capacity (about 8 kbps), without significant percep-tual  distortion(SNR  about  21dB)  and  provides  robustness  against  common  signal  processing  attacks  such  as  highpass  filtering,echo, resampling, amplifying MPEGcompression (MP3).etc.


Simulation and Optimization of the Effect of Work Function on Electrical Characteristics of Nano-MOSFETs
[Full Text]

Ooi Chek Yee, Aissa Boudjella and Lim Soo King

Numerical  simulations  have  been  performed  to  investigate  the  electronic  transport  properties  which  cause  drain
current to flow through the Silicon (Si) channel of double-gate (DG) fully-depleted (FD) n-MOSFET at nanoscaled dimension. In this investigation, the top and bottom gate contact work function is varied from 4.45 eV to 4.75 eV. These work function ranges cover  some  arbitrary  transition  metal  elements.  The  simulation  is  carried  out  at  room  temperature.  Three  electron  transport models  will  be  presented  :-  (1)  ballistic  transport  using  Green’s  function  approach,  (2)  ballistic  transport  using  semiclassical approach,  and  (3)  drift  diffusion  transport.  Electrical  data  such  as  the  subbands  energy  profile,  2D  electron  density,  current-voltage (I-V) and average electron velocity graphs have been plotted to study the effects of variations in the top and bottom gate contact work function. The effect on I-V characteristics are analyzed in terms of leakage current, on-state current and threshold


Integration of Expert systems with Intelligent Agents
[Full Text]

P. Venkata Subba Reddy

Software  Engineering is difficult with Artificial Intelligence particularly Expert Systems. The Intelligent Agents and
Expert Systems has common features like search and actions(reasoning). Software Engineering shall be made easy by Inte-grating  Expert  Systems  with  Intelligent  Systems    in  this  paper  integrated  Expert  Systems  with  Intelligent Agents  for  Software Engineering because  Expert Systems and  Intelligent Agents. are having common factors like search engine,  reasoning and user interface.  Infrastructure for implementation of Intelligent Agents/Expert Systems is discussed.


A Web Platform for Collaborative Multimedia Content Authoring Exploiting Keyword Search Engine and Data Cloud [Full Text]

Sonia Bergamaschi, Matteo Interlandi and Maurizio Vincini

The composition of multimedia presentations is a time- and resource-consuming task if not afforded in a well-defined manner. This is particularly true when people having different roles and following different high-level directives,
collaborate in the authoring and assembling of a final product. For this reason we adopt the Select, Assemble, Transform and Present (SATP) approach to coordinate the presentation authoring and a tag cloud-based search engine in order to help users in efficiently retrieving useful assets. In this paper we present MediaPresenter, the framework we developed to support companies in the creation of multimedia communication means, providing an instrument that users can exploit every time new communication channels have to be created.