Home
Call For Papers
Submission
Author
Registration
Publications
About
Contact Us

  Speaker Dependent Word Recognition Using MFCC and VQ  
  Authors : Nitin N Lokhande; Bhagsen J Parvat; Chandrakant Kadu
  Cite as:

 

The paper present effective method for recognition of digit, numbers. Most of speech recognition systems contain two main modules as follow “feature extraction” and “feature matching”. In this project, (MFCC) Mel Frequency Cepstrum coefficient algorithm is used to simulate feature extraction module. Using this algorithm, the Cepstral Coefficients are calculated on Mel frequency scale. VQ (vector quantization) method will be used for reduction of amount of data to decrease computation time. In the feature matching stage Euclidean distance is applied as similarity criterion. Because of high accuracy of used algorithms, the accuracy of this speech recognition system is high.

 

Published In : IJCSN Journal Volume 4, Issue 2

Date of Publication : April 2015

Pages : 420 - 425

Figures : 08

Tables : --

Publication Link : Speaker Dependent Word Recognition Using MFCC and VQ

 

 

 

N N Lokhande : working as Assistant Professor in Instrumentation and Control Engineering department of Pravara Rural Engineering College, Loni Maharashtra since last 08 years. He has completed Master of Engineering in Process Instrumentation. His field of Interest is signal processing and control systems.

B.J. Parvat : working as Associate Professor in Instrumentation and Control Engineering department of Pravara Rural Engineering College, Loni Maharashtra since last 14 years. He has completed Master of Technology in Process Instrumentation. He is pursuing PhD at SGGSI&T Nanded. His field of Interest is process control and control systems.

C.B.Kadu : working as Associate Professor in Instrumentation and Control Engineering department of Pravara Rural Engineering College, Loni Maharashtra since last 15 years. He has completed Master of Engineering in Process Instrumentation. He is pursuing PhD at COEP Pune. His field of Interest is process control and control systems.

 

 

 

 

 

 

 

Mel frequency Cepstral coefficient

Speech Recognition

Voice Activity Detection

Vector Quantization

This paper presents the speaker dependent digit recognition system using MFCC feature extraction algorithm and VQ as classification algorithm. Results are obtained on English database with codebook size32 and 64, recognition results are 86.26% and 100% respectively. Number of centroids increases the recognition rate also increases.

 

 

 

 

 

 

 

 

 

[1] John R. Deller, Jr., John H. L. Hansen, John G. Proakis, Discrete-Time Processing Of Speech Signals, John Wiley & Sons, inc., publication, IEEE Press. [2] Lawrence Rabiner and Biing-Hwang Juang, Fundamentals of speech Recognition, Prentice Hall, Englewood Cliffs, N.J., 1993. [3] Mikael Nilsson, Marcus Ejnarsson. “Speech Recognition using Hidden Markov Model”. Department of Telecommunications and Speech Processing, Blekinge Institute of Technology. 2002. [4] L.R. Rabiner and R.W. Schafer, Digital Processing of Speech Signals. Englewood Cliffs, NJ: Prentice-Hall, (1978). [5] R. M. Gray, ``Vector Quantization,'' IEEE ASSP Magazine, pp. 4--29, April 1984. [6] M.A.Anusuya, S.K.Katti Speech Recognition by Machine: A Review, (IJCSIS) International Journal of Computer Science and Information Security, Vol. 6, No. 3, 2009. [7] Junqua J-C., Haton J-P., (1996), “Robustness in ASR: Fundamentals and Applications”, Kluwer Academic Publishers.