Volume : V, Issue : VI, June - 2015

Vocabulary Based Digit Recognition

Dr. Anuj Kumar, Divya Rashmi

Abstract :

 A feed forward multilayer neural network is trained by Back propagation method for speaker independent isolated word recognition. Mel Frequency Cepstral Coefficients (MFCC) are extracted as speech features. These features are used to train the Multi Layer Feed Forward network(MLFFN) Network. The same routine is applied to signals during recognition stage and unknown test patterns are classified to the nearest pattern. This paper discusses the recognition of Hindi digits based on small vocabulary. Analysis based on varying number of hidden neurons in the network is presented here. The environment and is tested against data in test database created in similar environment. It has been observed that MLFFN works as good classifier for text data. For experimental purpose number of features extracted was changed and it has been observed that number of speech features extracted of isolated plays a very important rule in recognition of isolated Hindi digits through machine.

Keywords :

Article: Download PDF   DOI : 10.36106/ijar  

Cite This Article:

DR. ANUJ KUMAR, DIVYA RASHMI Vocabulary Based Digit Recognition Indian Journal of Applied Research, Vol.5, Issue : 6 June 2015

Number of Downloads : 667

References :