Speech Based Access for Agricultural Commodity Prices in Tamil

Author(s):

S. Yogapriya , Saranathan College of Engineering; Dr. P. Shanmugapriya, Saranathan College of Engineering; Dr. V. Mohan, Saranathan College of Engineering

Keywords:

GMM-UBM, i-vector, MFCC, Speech Recognition

Abstract:

Machine-Human interaction plays an important role for many users in order to access the information. The interactions could be in the combination of text, speech, facial expressions etc., or separately. The objective of the proposed work is to develop a speech based agricultural commodity prices accessing system in which recognition of user queries is implemented using i-vector approach. In the proposed work, i-vector based approach is adopted for speech recognition. The recognition tasks can be conducted for both speaker dependent and speaker independent case. The experimental results show that the Equal Error rate (EER) obtained for the proposed system is improved and the performance of the system is improved when increasing the number of Gaussian mixture components and also number of i-vector components.

Other Details:

Manuscript Id	:	IJSTEV4I7040
Published in	:	Volume : 4, Issue : 7
Publication Date	:	01/02/2018
Page(s)	:	75-81

Download Article