The human voice recognition algorithm in determining gender classification

Abstract

This study aims to determine the type of person's voice to group them based on the type of voice. So far, determining the type of person's voice manually requires an expert such as a vocal trainer. To determine only one or two persons' voice will be quick, but to determine more person's voices such as a choir group will be difficult because it takes time and costs. This manual method is not effective to determine the type of sound in a large group. Therefore, we need a system to determine the type of voices. This study use Mel Frequency Cepstral Coefficient (MFCC) and K-Nearest Neighbor (K-NN). MFCC extracts voice features and K-NN classifies voice types.