How to use a trained neural network in GUIDE?

I trained a convolutional neural network using the pre-trained network ResNet50 for medical image classification. Now I want to create a user interface in GUIDE where a user can upload a picture of a skin tumor and the net will classify it and return the label for the image. My problem is I'm not sure where should I put the network in the interface code. I am using axes for the image and a push button for the upload function, and I was thinking of using another push button for the classification.

Answer 1

[Prasanna]

Hi Malina,

To create a UI in GUIDE and integrate the pre-trained network with the same: You can load the network, set up the GUI to handle image uploads, and classify the uploaded images upon a button press. Some further instructions on the above steps include:

• Load the Pre-trained Network: Load your pre-trained network when the GUI initializes. This can be done in the “OpeningFcn” of your GUIDE-generated .m file, which is executed before the GUI becomes visible.
• Upload and Display the Image: In the callback for your upload button, you can implement functionality to upload and display the image within the axes you've designed.
• Classify the Uploaded Image: For the classification button, you can set up a callback that processes the uploaded image according to the input requirements of your pre-trained network (e.g., resizing), performs classification, and then displays or returns the classification result.

You can refer the following documentation for more information on writing callbacks in GUIDE:

• https://www.mathworks.com/help/matlab/creating_guis/write-callbacks-using-the-guide-workflow.html
• https://www.mathworks.com/help/matlab/creating_guis/share-data-among-callbacks.html

Hope this helps.