User Voice management and Power spectrum analysis for Voice Recognition Systems

  • M.Anil Kumar et. al


ABSTRACT: This manuscript deals with the usage of an ongoing custom/user input voice acknowledgment framework utilizing MATLAB (R2016a) to verify a given client’s framework with the goal of client’s accessibilities may get to it. Voice acknowledgment frameworks have two stages, preparing and testing. Amid the preparation stage, the trademark highlights of the speeches are either classified into male or female voices in databases, for each such scenario we tend to provide stage vise operations for the design system. To improve the capabilities, we provide TEST and TRAIN framework. In the testing stage, the client’s voice highlights testing voice contrasted by a tests voice in decibels and decides whether similarity is present or not. In proposed examination, MFCC’s is picked for speaking to an element vectors to client’s audio that it’s precisely reproduces an conduct to hearing voice. Its normal for the MFCCs creates it a phenomenal proportion as audio attributes. A component coordinating procedure later presented by oppressing an MFCCs {DTW and MED approaches} to changed VQ. In such situations, noise is a main parameter of consideration that antagonistically impacts a voice acknowledgment framework. The paper tends to this issue by using spectral subtraction to eradicate noises which improvises in ecological speech signal flag due this scenario we could expand the vigor of the framework.

How to Cite
et. al, M. K. (2020). User Voice management and Power spectrum analysis for Voice Recognition Systems. International Journal of Advanced Science and Technology, 29(2), 2325 - 2333. Retrieved from