Randomly splitting Face (Image) Dataset

Randomly splitting Face (Image) Dataset into training and testing sets

This Matlab function can be used to randomly split any given face (image) dataset into training and testing sets. The function considers class/person labels. For example, in ORL Face dataset, there are 40 persons each with 10 images; the function can randomly split the 10 images of each person into 5 train images and 5 test images, as follows:

load ORL_FaceDataSet % uploaded here
[TrainSet TestSet]=Split_DataSet(ORL_FaceDataSet, 10, 5, 5);

Note that, you can select different numbers of train and test images.