How do you get the Classification Layer Probabilities for pretrained CNNs ?

Hello,

I have code that runs through a file of images and classifies each, using the pretrained Squeezenet network. I want to get the results of the classification layer and save all the probabilities and the end classification choice. How do I do this?

Thank you!

for k = 1:numFiles
try
    full_path=fullfile(imgs(k).folder, imgs(k).name);
    mydata{k} = imread(full_path);
    i = imresize(mydata{k},[227,227]);
    a = classify(net, i)
    
   imshow(i); title(char(a))
catch ME
    fprintf("The imread call threw the following error:\n%s\n", ME.message)
    fprintf("The error identifier for this error is:\n%s\n", ME.identifier)
end
    
end

Answer 1

[Divya Gaddipati]

You can use activations to get the output at a certain layer

out = activations(net,X,'layerName');

For more information, refer to the link below:

https://www.mathworks.com/help/deeplearning/ref/seriesnetwork.activations.html

[AK]

How do i output the activation information once i have it? Like if i want to print it ? Thanks!