Informatics Publishing Limited

K. Sweta ¹, B. Amutha ¹

Affiliations
1 Department of Computer Science, SRM University, IN

Abstract

The conversion of speech to text for the regional language, TAMIL is being done using HMM (Hidden Markov Model). As the speech to text conversion is yet to be exploited in Tamil language, the purpose of converting speech to text for Tamil language becomes essential. Tamil is a highly inflectional language. The primary aim of this research is to understand the Tamil spoken word and convert it into Tamil text. The full-fledged conversion of all Tamil words through speech to text may be performed in the near future. In the first trial, only 25 familiar Tamil words have been identified, recognized through English, converted and displayed in Tamil text. This conversion of speech to text is done with the help of HMM (Hidden Markov Model) process for 25 Tamil words and the accuracy of conversion is found to be 83%.

Keywords

HMM, Speech to Text Conversion, VAD, Markov Process, Azhagi.

Full Text

     

V. Swarna Priya ¹, B. Amutha ¹

Affiliations
1 Department of Computer Science, SRM University, Kattankulattur, Chennai, IN

Abstract

This research is based on the conversion of English speech into English text with an efficient system for independent speaker speech recognition based on Neural Network Approach using Viterbi Algorithm. To recognize the English words consider all the accents of same spoken word, so that matching process with the actual word does not lead to any difficulties. There are 26 characters in English. It is well known that the pronunciation of a word depends heavily on the background. Speech dependent & speaker independent technique can be used and English words must be recognized.
HMM n to 1 encoder and HMM 1 to n decoder for finding the speech to text with the help of viterbi algorithm. There may be two or more pronunciation for the same word so the database should be maintained in that all expected accent should be there for a particular word. The main focus is to clearly get the spoken word for the different pronunciation/accent. This approach is better than others as the identification becomes easier.

Keywords

Viterbi Algorithm, HMM, Speech to Text Conversion, English WORDS.

Full Text

     

K. Deeba ¹, B. Amutha ¹

Affiliations
1 Department of Computer Science and Engineering, SRM University, Kattankulathur, Chennai - 603203, Tamil Nadu, IN

Abstract

Objectives: To make a comparative study about different classification techniques of data mining. Methods: In this paper some data mining techniques like Decision tree algorithm, Bayesian network model, Naive Bayes method, Support Vector Machine and K-Nearest neighbour classifier were discussed. Findings: Each algorithm has its own advantages and disadvantages. Decision tree technique do not perform well if the data have smooth boundaries. The Naive Bayesian classifier works with both continuous and discrete attributes and works well for real time problems. This method is very fast and highly scalable. The drawback of this technique is when a data set which has strong dependency among the attribute is considered then this method gives poor performance. KNN can perform well in many situations and it particularly suits well for multi-model classes as well as applications in which an object can have many labels. The drawback of KNN is it involves lot of computation and when the size of training set taken is large then the process will become slow. Support vector machine suites well when the data need to be classified into two groups. Application: This paper is to provide a wide range of idea about different classification algorithms..

Keywords

Bayesian Network, Data Mining, Decision Tree, K-Nearest Neighbour Classifier, Naive Bayes, Support Vector Machine.

Full Text

   

B. Amutha ¹, A. D. Giridhar ¹, Siddharth Sahani ¹

Affiliations
1 Department of CSE, SRM University, Chennai - 603203, Tamil Nadu, IN

Abstract

Objective: Passive measures of memory are in increasing demand since active measures like EEG, fMRI etc. are not a viable way of cognitive measurement in environments where such devices are difficult to be accessed or for age groups and set of people to which such measures are difficult. Methods: Thus, in this paper, we propose a novel mathematical model for a passive measure of recall memory in children of age group 10 to 15 who are suffering from memory difficulties owing to affective effects. Findings and Applications: Though such passive measures are still abstract, such a methodology can be used ‘on the go’ and such can be supported with the help of clinical methods that support the improvement of cognitive abilities like memory.

Keywords

Affect, LTM, Memory, Memory Disorders, Memory Loss Causes, Reaction Time, Recall, Recognition, STM, Types of Affect.

Full Text

   

A. S. Suryapattatharasi , B. Amutha , D. Venkatesan

Abstract

Generally, the unicast and the multicast routing protocol are used to deliver messages between randomly chosen nodes. The major problem with these type of routing is that they do not process the distributed information's. Several algorithms or protocols such as shortest routing protocol and position based protocol are found to ensure effective routing in sparse network. These techniques focus on delivering the messages to certain nodes. However these techniques or protocols limit the process of retaining the information regarding the missing nodes. The proposed boomerang protocol effectively receives the messages from the nodes at any geographical area. The boomerang protocol attempts to utilize the physical location based information to increase the performance of geo cast delivery. A unique feature of this protocol is that it records the geographical trajectory while moving away from the origin, thus paving a way to exploits the recorded trajectory in case of returning back. This protocol simply attempts to provide more efficient routing method while maintaining the information of the missing node. Boomerang protocol finds a wide application in LAN, WAN, and other mobile communication network.

Keywords

Geo Cache, GPS, Mobile, Location Aware Information, Infrastructure Less Data Management

Full Text