confusion matrix for image retrieval

im = fullfile(pn, fn);

images_query = imageDatastore(rootFolder, 'IncludeSubfolders',true, 'LabelSource','foldernames'); %%'ReadFcn', @readCBIR

R = imread(im); % Read image

Input_Layer_Size_q = net.Layers(1).InputSize(1:2); % (1:2 = 1st 2 elemnts of input size), input layer size stored in this variable (Input_layer_size)

Resized_Test_image_q = augmentedImageDatastore(Input_Layer_Size_q, R, 'ColorPreprocessing','gray2rgb'); %% For defining test image replace "Testing _image with test folder

%Extract feature

train_feature = activations(net, Resized_Training_image, 'Animal Feature Learner', 'OutputAs', 'Rows');

query_feature = activations(net, Resized_Test_image_q, 'Animal Feature Learner', 'OutputAs', 'Rows');

%Equation 2

a = query_feature; % transposing

b = transpose(1-a);

%Equation 3

c = zeros(Number_of_Classes,Number_of_Training_images);

d = sqrt(sum((query_feature' - train_feature') .^ 2)); % other method eucledian: giving all images from same category maybe something is wrong

for e = 1 : Number_of_Training_images

f = b.*d(:,e);

c(:, e) = f;

end

c = sqrt(sum(c))';

% Fetch top 25 similar images

g = sort(c);

[~, n] = sort(c);

n = n(1:25);

files = cell(1, 25);

for h =1:25

files{h} = Training_image.Files{n(h)};

end

%Display query image

figure;

imshow(R);

title('query')

% Display retrived images

figure;

montage(files);

title("retrived")

figure;

confusionchart(Testing_image.Labels, Predicted_Label,'Normalization','row-normalized', 'Title', 'Normalised ConfMat'); % normalised values for different rows, col, etc

confMat = confusionmat(Testing_image.Labels, Predicted_Label); %confusionmat generates cofusion matrix

How do I find confusion matrix, precison, RMS error for image retrival?

Answer 1

[Omega]

Hi,

I understand that you would like to calculate precision, RMSE(Root mean square error) and plot the confusion matrix for your image retrieval task.

To find the confusion matrix, precision, and RMSE for image retrieval, you need to have the ground truth labels of the test images and the predicted labels obtained from the retrieval process.

1. Confusion Matrix:

Assuming you have the predicted labels “Predicted_Label” and the true labels “Testing_image.Labels” you can use the “confusionmat” function to generate the confusion matrix and plot it using the “confustionchart” function in MATLAB. Here's an example:

confMat = confusionmat(Testing_image.Labels, Predicted_Label); 
confusionchart(Testing_image.Labels, Predicted_Label,'Normalization','row-normalized', 'Title', 'Normalised ConfMat'); 

2. Precision:

Precision can be calculated as the number of true positives divided by the sum of true positives and false positives for each class. You can use the confusion matrix to calculate precision. Here's an example for a specific class:

classIndex = 1; % Index of the class for which you want to calculate precision 
truePositives = confMat(classIndex, classIndex); 
falsePositives = sum(confMat(:, classIndex)) - truePositives; 
precision = truePositives / (truePositives + falsePositives);

Note that, precision is typically calculated per class, so you may need to loop through the classes to obtain the overall precision value.

3. RMSE:

RMSE (Root Mean Square Error) is not typically used for evaluating image retrieval tasks. RMSE is a metric commonly used in regression tasks to measure the average squared difference between predicted and actual values. Still, if you would like to calculate RMSE you can do so by using MATLAB’s “rmse” function.

You can refer to the following documentation links which might help you in understanding the functions better:

• confusionmat: https://www.mathworks.com/help/stats/confusionmat.html
• confusionchart: https://www.mathworks.com/help/stats/confusionchart.html
• rmse: https://www.mathworks.com/help/matlab/ref/rmse.html