Comparison of Different Normalization Techniques on Speakers’ Gender Detection

Journal Title: Mehmet Akif Ersoy Üniversitesi Uygulamalı Bilimler Dergisi - Year 2018, Vol 2, Issue 2

Abstract

In this study, the effect of Short-time Mean and Variance Normalization (STMVN), Shorttime Cepstral Mean and Scale Normalization (STMSN), Min-Max Normalization, Z-Score Normalization and Standard Deviation Normalization techniques on the classification performance was investigated in determining speakers’ gender. In the study, voice records which belongs to 192 male and 192 female speakers from TIMIT data set were used as data set. Features were extracted from Mel Frequency Cepstral Coefficients (MFCC) technique by using voice records and extracted features’ dimension was reduced to Principal Component Analysis (PCA), then normalized with different techniques. Support Vector Machine (SVM) was used as classifier. As a result of study, it was observed that, the highest accuracy in speakers’ gender estimation is obtained as 98.18% from features which were normalized with Standard Deviation Normalization technique and other normalization techniques were reduced accuracy.

Authors and Affiliations

Serhat Celil İleri, Armağan Karabina, Erdal Kılıç

Keywords

Related Articles

The Investigation of the Effects of Different Filters on Mammogram Images

Mammography is a widespread imaging technique to early detect breast cancer. It can detect micro scale calcium deposits (microcalcification) known as early signs of breast cancer. The detection of microcalcifications on...

İkinci Derece Zaman Gecikmeli Sistemler için Analitik PD Denetleyici Tasarımı

Bu yayında ikinci derece zaman gecikmeli sistemlerin kararlılığı ve performansı için oransal türev denetleyicinin analitik tasarımı şeması gösterilmiştir. Bu sistem için Bode’nin ideal transfer fonksiyonunun karakteristi...

Gerontoloji Lisans Eğitiminde Veri Araştırmaları Yoluyla İstatistik Öğretimi

National and international reports, as well as independent studies showed that the ratio of elderly in population tends to grow fast in near future, and the average life expectancy has been increasing for almost all coun...

Social Capital: An Assessment for Turkey

Considering the fact that social capital is an important area for practitioners and policymakers who want to develop a more productive and inclusive economic structure, the reason for the interest in social capital, the...

Kurs Oranı ve Artık Gaz Kesrinin Otto Çevrimli Bir Motorun Performansına Etkisi

The finite time thermodynamics (FTT) model is often used to study the design parameters of internal combustion engines in a practical way. In this study, the effects of stroke ratio and residual gas fraction on the perfo...

Download PDF file
  • EP ID EP418309
  • DOI 10.31200/makuubd.410625
  • Views 145
  • Downloads 0

How To Cite

Serhat Celil İleri, Armağan Karabina, Erdal Kılıç (2018). Comparison of Different Normalization Techniques on Speakers’ Gender Detection. Mehmet Akif Ersoy Üniversitesi Uygulamalı Bilimler Dergisi, 2(2), 1-12. https://europub.co.uk/articles/-A-418309