cnn_imagenet_minimal.m

function cnn_imagenet_minimal()
% CNN_IMAGENET_MINIMAL   Minimalistic demonstration of how to run an ImageNet CNN model

% Setup MatConvNet.
run(fullfile(fileparts(mfilename('fullpath')), ...
  '..', '..', 'matlab', 'vl_setupnn.m')) ;

% Download a pre-trained CNN from the web.
if ~exist('imagenet-vgg-f.mat', 'file')
  fprintf('Downloading the VGG-F model ... this may take a while\n') ;
  urlwrite('http://www.vlfeat.org/matconvnet/models/imagenet-vgg-f.mat', ...
    'imagenet-vgg-f.mat') ;
end

% Load the model and upgrade it to MatConvNet current version.
net = load('imagenet-vgg-f.mat') ;
net = vl_simplenn_tidy(net) ;

% Obtain and preprocess an image.
im = imread('peppers.png') ;
im_ = single(im) ; % note: 255 range
im_ = imresize(im_, net.meta.normalization.imageSize(1:2)) ;
im_ = im_ - net.meta.normalization.averageImage ;

% Run the CNN.
res = vl_simplenn(net, im_) ;

% Show the classification result.
scores = squeeze(gather(res(end).x)) ;
[bestScore, best] = max(scores) ;
figure(1) ; clf ; imagesc(im) ;
title(sprintf('%s (%d), score %.3f',...
   net.meta.classes.description{best}, best, bestScore)) ;