Mercurial > hg > camir-aes2014
diff toolboxes/FullBNT-1.0.7/netlab3.3/mdn2gmm.m @ 0:e9a9cd732c1e tip
first hg version after svn
author | wolffd |
---|---|
date | Tue, 10 Feb 2015 15:05:51 +0000 |
parents | |
children |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/toolboxes/FullBNT-1.0.7/netlab3.3/mdn2gmm.m Tue Feb 10 15:05:51 2015 +0000 @@ -0,0 +1,41 @@ +function gmmmixes = mdn2gmm(mdnmixes) +%MDN2GMM Converts an MDN mixture data structure to array of GMMs. +% +% Description +% GMMMIXES = MDN2GMM(MDNMIXES) takes an MDN mixture data structure +% MDNMIXES containing three matrices (for priors, centres and +% variances) where each row represents the corresponding parameter +% values for a different mixture model and creates an array of GMMs. +% These can then be used with the standard Netlab Gaussian mixture +% model functions. +% +% See also +% GMM, MDN, MDNFWD +% + +% Copyright (c) Ian T Nabney (1996-2001) +% David J Evans (1998) + +% Check argument for consistency +errstring = consist(mdnmixes, 'mdnmixes'); +if ~isempty(errstring) + error(errstring); +end + +nmixes = size(mdnmixes.centres, 1); +% Construct ndata structures containing the mixture model information. +% First allocate the memory. +tempmix = gmm(mdnmixes.dim_target, mdnmixes.ncentres, 'spherical'); +f = fieldnames(tempmix); +gmmmixes = cell(size(f, 1), 1, nmixes); +gmmmixes = cell2struct(gmmmixes, f,1); + +% Then fill each structure in turn using gmmunpak. Assume that spherical +% covariance structure is used. +for i = 1:nmixes + centres = reshape(mdnmixes.centres(i, :), mdnmixes.dim_target, ... + mdnmixes.ncentres)'; + gmmmixes(i) = gmmunpak(tempmix, [mdnmixes.mixcoeffs(i,:), ... + centres(:)', mdnmixes.covars(i,:)]); +end +