Inception-ResNet-v2 is a pretrained model that has been trained on a subset of the ImageNet database. The model is trained on more than a million images, has 825 layers in total, and can classify images into 1000 object categories (e.g. keyboard, mouse, pencil, and many animals).
Opening the inceptionresnetv2.mlpkginstall file from your operating system or from within MATLAB will initiate the installation process for the release you have.
net = inceptionresnetv2()
% Read the image to classify
I = imread('peppers.png');
% Crop image to the input size of the network
sz = net.Layers(1).InputSize
I = I(1:sz(1), 1:sz(2), 1:sz(3));
% Classify the image using Inception-ResNet-v2
label = classify(net, I)
% Show the image and classification result
text(10, 20, char(label), 'Color', 'white' )
lgraph = layerGraph(net); % net.Layers -> net
this model gives me many size errors although the model scales the input data itself.
do you have some example for transfer learning this model ?
I am using this architecture for purely feature extraction (i.e. no testing). How do we subtract the average Image from the net to normalize our input images?
Hi Nicola, We are sorry for the inconvenience. The support for R2018a version will be made available very soon.
Hi, on matlab r2018a - windows server 2016, it's not possible to install it.
Add-On Manager show me a message: The support package is not compatible with your version of MATLAB or operating system.
Work only with r2017b?