How to load and predict using my trained deep neural network in MATLAB GUI window?

hi!

I have a trained deep neural network using my own dataset. How do i load my trained network in the GUI and predict the output using any random image. I have developed GUI using App Desinger. Please help

Tried to load network using different ways:

One way-

trained_net_vars = load('E:\Subject\newtrain_05_12.mat');

handles.neural_net = trained_net_vars.net;

guidata(hObject, handles)

Second way -

% [f, p, i]=uigetfile('*.mat');

% f_name=fullfile(p,f);

% a=load(f_name);

% A_cell = struct2cell(a);

Answer 1

[Sanjana]

Hi,

I understand that you're facing difficulties in creating a GUI window to load a pretrained model and perform predictions on random images, and then display the prediction. Here is a possible solution:

Firstly, you can add a “button” and an “Image” component to the canvas in the App Designer. Then, write a “callback” function for the “button” component as follows:

function LoadImageButtonPushed(app, event)
            net = load('modelName.mat');
            [filename, pathname] = uigetfile({'*.jpg;*.png;*.bmp', 'Image files'}, 'Select an image');
            filepath = fullfile(pathname, filename);
            app.Image.ImageSource = filepath;
            img = imread(filepath);
            prediction = predict(net,img);
            app.OutputLabel.Text = prediction;
 
        end

Hope this helps!

[shazia]

hello , if i want to pass my test data (not image) and the genrated trained net to GUI for prediction, how i would do that could you plz elobrate.

thank you