[Edu-sig] PYTHON for audio processing

[David Handy]

Hi Sundeep -
 
I don't know anything about the MFCC, DTW, or ANN algorithms you mentioned.
I would have to guess that they do some kind of transformation to frequency domain as a first stage, and probably use something like the FFT. The numpy (Numeric Python) package provides FFT and other accelerated numeric capabilities. You should take a look at it.
 
As far as HTK goes, if it is a library written in C/C++ with a C interface, supplied as a DLL or shared object library, and you have C programming experience, you could use the Python ctypes module and create your own Python interface to it that way, without having to write any C code. Just a thought.
 
David H
 
-----Original Message-----
From: "Sundeep Sivan" <sundeepsivan at gmail.com>
Sent: Tuesday, September 3, 2013 12:21pm
To: edu-sig at python.org
Subject: [Edu-sig] PYTHON for audio processing



Hi,
I am an M.TECH student. I would like to do my main project in python. and my main project is "Speaker recognition"(ie. recognizing persons from their voice). For feature extraction i would like to use MFCC(Mel frequency cepstral coefficients) and For feature matching i may use Hidden markov model or DTW(Dynamic time warping) or ANN. When i googled about it their are many toolkits like HTK etc avilable for MATLAB. But i didnt get any tools for python. So my question is that can i do my project in python? Also is their any toolkits or modules avilable for my project. Kindly help.
Regards
Sundeep
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/edu-sig/attachments/20130916/124425f8/attachment.html>