matlab-bow/compute_kernel_map.m at master · jmrr/matlab-bow · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
function [kernelTrain,kernelTest] = compute_kernel_map(dataset,datasetDir,dictDir,params,kernelType)

% Load dictionary of training images

dictFname = sprintf('dictionary_%d.mat',params.dictionarySize);
load(fullfile(dictDir,dictFname));

stackTrain = [];
stackTest  = [];

for cat = 1:length(dataset)

    % Initial parameters
    numImgs  = length(dataset(cat).files);
    catPath  = fullfile(datasetDir,dataset(cat).className);

    % Concatenate the svm training and test features
    trainIndices = dataset(cat).train_id;
    testIndices  = dataset(cat).test_id;

    numTrainSam  = sum(trainIndices);
    numTestSam   = sum(testIndices);
    % Load training features

    load(fullfile(datasetDir,dataset(cat).className,...
        [params.encodingMethod '_' num2str(params.dictionarySize) '.mat']));

    stackTrain = [stackTrain; HoVW_all(trainIndices,:)];
    stackTest  = [stackTest; HoVW_all(testIndices,:)];

end % end for categories

% Normalize

stackTrain = stackTrain./repmat(sum(stackTrain.^2,2)+eps,[1, params.dictionarySize]);
stackTest  = stackTest./repmat(sum(stackTest.^2,2)+eps,[1, params.dictionarySize]);

% Compute the feature maps
psiTrain = vl_homkermap(stackTrain',1,kernelType);
psiTest  = vl_homkermap(stackTest',1,kernelType);

% Compute the kernels
kernelTrain = psiTrain'*psiTrain;
kernelTest  = psiTest'*psiTest;

end % end