An Analysis of Speech Signal for Multiple Languages through Spectrum Analysis

Volume 3 Issue 4 December - February 2017

Research Paper

An Analysis of Speech Signal for Multiple Languages through Spectrum Analysis

Pooja Yadav*, Vinay Kumar Jain**

* PG Scholar, Electronics and Telecommunication Engineering, SSTC, Bhilai, India.

** Research Scholar, Swami Vivekananda Technical University, Bhilai, India.

Yadav, P., and Jain, V. K. (2017). An Analysis of Speech Signal for Multiple Languages through Spectrum Analysis. i-manager’s Journal on Pattern Recognition, 3(4), 22-27. https://doi.org/10.26634/jpr.3.4.13540

Abstract

The listeners outperform Automatic speech recognition structures in each and every speech reputation task. Modern excessive-tech automated speech recognition systems carry out very well in environments, wherein the speech indicators are reasonably easy. In maximum of the instances, popularity with the aid of machines degrades dramatically with mild adjustment in speech signals or talking environment, for this reason complicated algorithms are used to symbolize this unpredictability. So, the speech can be easily identified. Speech generation gives many possibilities for private identity, which is practical and non-intrusive. Besides that, speech era offers the capability to verify the identity of a person remotely over long distance by using an ordinary phone. In this paper, the authors have proposed a technique to apprehend any words or speech thru the spectrogram analysis. This technique is used to look at the ideas of speaker reputation in multiple languages and apprehend its uses in identification and verification systems and to assess the recognition capability of various voice functions and parameters. To find out the technique, this is appropriate for Automatic Speaker Recognition systems in phrases of reliability and computational efficiency.