High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

A Review Paper on Voiced/Unvoiced Classification

Author(s):

Bhumika Nirmalkar , Rungta College of Engineering and Technology Bhilai, India; Bhumika Nirmalkar, Rungta College of Engineering and Technology Bhilai, India; Dr. Sandeep Kumar, Rungta College of Engineering and Technology Bhilai, India

Keywords:

Voiced/unvoiced classification, Cepstrum, EMD

Abstract

This paper presents voiced/unvoiced classification based on various methods like empirical mode decomposition, wavelet, cepstrum, zero crossing rate, short time energy etc and statistical model such as neural network, hidden Markov model (HMM) and Gaussian mixture model (GMM).In most of these techniques the voiced/unvoiced classification is usually perform by means of placing a threshold value with a few acoustic functions say, short time energy, zero crossing rate and so on. The primary trouble is the purpose of effective threshold which impact the type issues performance is likewise comes to a decision the choice of threshold. We have observed that a hybrid technique for the voiced/unvoiced classification can enhance the performance of the existing schemes.

Other Details

Paper ID: IJSRDV4I40986
Published in: Volume : 4, Issue : 4
Publication Date: 01/07/2016
Page(s): 1065-1070

Article Preview

Download Article