5
$\begingroup$

I would like to use MFCC analysis to distinguish between a male voice and a female voice.

Are there any particular quantifiable characteristics consistent with, and exclusive to, either gender? (E.g., energy, fundamental frequency, excitation, the presence of certain waveforms, etc.)

Improve this question
$\endgroup$
2
  • 2
    $\begingroup$ That, for one is a research question. I was asked the same elsewhere and people are currently working on it. $\endgroup$
    – Naresh
    Commented Dec 13, 2012 at 9:45
  • $\begingroup$ What accuracy or error level does your analysis require? $\endgroup$
    – hotpaw2
    Commented Dec 13, 2012 at 15:12

1 Answer 1

Reset to default
6
$\begingroup$

Main difference is the frequency of the fundamental, which is about an octave higher for women with the split point around 160 Hz or so. A fundamental lower than 160 Hz is most likely a male and a fundamental higher is most likely a female. A good overview over a number of studies on the topic can be found at http://www2.ling.su.se/staff/hartmut/f0_m%26f.pdf

Improve this answer
$\endgroup$
2
  • 2
    $\begingroup$ The following immediate question would be: "given a high-pitch male voice and a low-pitch female voice with roughly the same fundamental frequency, which OTHER features/parameters could be used to identify each voice's gender?" $\endgroup$
    – heltonbiker
    Commented Dec 13, 2012 at 18:45
  • $\begingroup$ Formants position. Given that pitch is somehow captured by the MFCCs (order 1 MFCC is equivalent to a spectral centroid, and for a given phoneme will follow f0) ; and that many pitch estimators are prone to octave errors, using the f0 as a feature for gender classification is risky... $\endgroup$
    – pichenettes
    Commented Dec 13, 2012 at 19:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.