Code covered by the BSD License  

Highlights from
Useful Matlab Functions for Speaker Recognition Using Adapted Gaussian Mixture Model

Useful Matlab Functions for Speaker Recognition Using Adapted Gaussian Mixture Model

by

 

05 Jun 2011 (Updated )

This submission includes useful MATLAB functions for speaker recognition using adapted GMM.

index=gmmprob_index(mix,data,N)
function index=gmmprob_index(mix,data,N)
% This code finds N-top Gaussians of UBM for a feature matrix.
% Written by Md. Sahidullah (Graduate Student, IIT Kharagpur)
% For any sort of queries or suggestions please mail to 
%                                                    sahidullahmd@gmail.com
%--------------------------------------------------------------------------
%Usage Input: 
%        : INDEX=gmmprob_index(mix,data,N)
%        : mix=Netlab GMM structure of UBM
%        : data=input data ( Txd matrix where d = mix.nin)
%        : N=No of Gaussian to be selected
%      Output:
%        : index: a TxN matrix of Gaussian indices
% Use this with Netlab toolbox
% http://www1.aston.ac.uk/eas/research/groups/ncrg/resources/netlab/
% If you have any query or suggestion please mail me sahidullahmd@gmail.com
b=mix.priors;
%-----------------Computation of Probability of all Gaussians--------------
a = gmmactiv(mix, data);
c=(a.*repmat(b,size(data,1),1))';
%----------------Sort Listing of Gaussians---------------------------------
[max_v max_p]=sort(c,'descend');
max_p=max_p';
%---------------Selction of Top-N modes------------------------------------
index=max_p(:,1:N);
%-----------------------End of Code----------------------------------------

Contact us