资源简介
PLDA模型训练matlab代码
代码片段和文件信息
%******************** Disclaimer *****************************************
% This program is distributed in the hope that it will be useful but
% WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
%
% The program is free for academic use. If you are interested in using the
% software for commercial purposes. Please contact
% Dr. Simon J. D. Prince
% Email: s.prince@cs.ucl.ac.uk
%
% Plase use the citation provided below if it is useful to your research:
%
% S.J.D. Prince and J.H. Elder 揚robabilistic linear discriminant analysis
% for inferences about identity? ICCV 2007.
%
% P. Li and S.J.D. Prince 揚robabilistic Methods for Face Registration and
% Recognition? In Advances in Face Image Analysis: Techniques and
% Technologies Y. Zhang (eds.) (in press).
%
%**************************************************************************
% function LogLikelihood = PLDA_Identification(ModelDataGalleryDataProbe)
%
% Face identification using PLDA based on model comparison
% Calculate the loglikelihoods of models that DataProbe match DataGallery
% using learned PLDA
%
% Inputs:
% Model - Learned PLDA model
% DataGallery - Gallery data: NFeature x nGallery
% DataProbe - Probe data: NFeature x nProbe
% Outputs:
% LogLikelihood - nGallery x nProbe matrix of loglikelihoods that
% assumes a probe image matches a gallery image
%**************************************************************************
% Date: 04-03-2010
function LogLikelihood = PLDA_Identification(Model DataGallery DataProbe)
% Get the learned model parameters
F = Model.F;
G = Model.G;
Sigma = Model.Sigma;
meanVec = Model.meanVec;
% Define space for log likelihoods
nGallery = size(DataGallery 2);
nProbe = size(DataProbe 2);
LogLikelihood = zeros(nGallery nProbe);
% Subtract mean from gallery and probe data
DataGallery = DataGallery - repmat(meanVec1nGallery);
DataProbe = DataProbe - repmat(meanVec1nProbe);
% Preprocess PLDA model and data to accelerate the computation
HIGHEST_N = 2; % highest number of faces explained by one variable
factorModel = preProcessPLDAModel(F G Sigma HIGHEST_N);
galleryDataPP = preProcessPLDAData(factorModel DataGallery);
probeDataPP = preProcessPLDAData(factorModel DataProbe);
% Precalculate single probability terms - senario that each gallery image
% is from a different identity
logLikeGalleryNoMatch = zeros(nGallery 1);
for cImage = 1:nGallery
logLikeGalleryNoMatch(cImage) = getLogLikeMatchPLDA(...
factorModel galleryDataPP(cImage));
end
% Loop over each probe
for cProbe = 1 : nProbe
% Loop over each gallery
for cGallery = 1 : nGallery
% Loglikelihood of this model w.r.t. the non-match terms
logLikeModel = sum(logLikeGalleryNoMatch(:))...
- logLikeGalleryNoMatch(cG
属性 大小 日期 时间 名称
----------- --------- ---------- ----- ----
文件 10537008 2010-01-05 12:04 Data.mat
文件 14638107 2010-03-04 15:11 PLDAModel.mat
文件 6762 2010-03-04 14:56 PLDA_Identification.m
文件 2180 2010-03-04 14:55 PLDA_Identification_Demo.m
文件 2764 2010-03-04 14:54 PLDA_Identification_Demo2.m
文件 8946 2010-03-04 15:07 PLDA_Train.m
文件 1815 2010-03-04 15:06 PLDA_Train_Demo.m
文件 5715 2010-03-04 15:14 PLDA_Verification.m
文件 2863 2010-03-04 15:32 PLDA_Verification_Demo.m
文件 4299 2010-03-04 15:35 Rand_Feature_Sampling_Demo.m
文件 3450 2010-03-04 15:41 Readme.txt
相关资源
- hogcascade训练,内有人头训练样本和使
- dsp开发matlab代码直接生成
- 数字调制的MATLAB代码 ASK FSK OOSK PSK
-
《基于MATLAB/Simuli
nk的系统仿真技术 - 手写数字识别Matlab代码
- 统计学习方法笔记部分算法含matlab代
- 《模式识别与智能计算:MATLAB技术实
- 音频检索的matlab代码
- 机器学习与MATLAB代码
- 深度学习工具包matlab代码
-
薛定宇 陈阳泉基于 MATLAB+Simuli
nk 的 - BLS宽度学习matlab代码.zip
- 单通道盲源分离SSA-ICA算法Matlab代码
- matlab代码实现模拟ps软件部分功能
- 深度置信神经网络的matlab代码.rar
- lenet_5模型-MATLAB代码实现
- LDA详解及其matlab代码
- Softmax回归模型(matlab代码)
- (全文) 自适应盲均衡技术 郭业才著
- 合成孔径雷达成像算法与实现.zip+ma
- 统计信号处理基础之实用算法开发(
- 元胞自动机matlab程序代码
- 经典手眼标定算法matlab代码
- (全文) 自适应盲均衡技术 郭业才著
- 深度置信神经网络的matlab代码
- 帧差法实现目标跟踪的Matlab代码
- 基于图像块低秩性的修复matlab代码
- 最先提出深度学习算法hinton的自动编
- 作业车间调度问题文档 和matlab代码
- Rectangling Panoramic Images via Warping论文的
评论
共有 条评论