Computer Vision News - June 2016

We will demonstrate two uses of TensorLab: as an exploratory tool and as a classification tool. For those, we will use the Yale face database . The database consists of 165 grayscale images in PNG format of 15 individuals. There are 11 images per subject, one per different facial expression or configuration: center- light, w/glasses, happy, left-light, w/no glasses, normal, right-light, sad, sleepy, surprised and wink. This database is publicly available and can be downloaded here . We will start by loading the database with the following code: Line 1: read the list of files under the ‘pathToYealFaceDB’ folder (you will need to set the ‘pathToYealFaceDB’ variable to that location in which you have downloaded the database). Line 2: initiate a zero matrix (tensor in our case) of the size 243x320x165 to hold the 165 images, the loop in lines 3-7 read the images one after the other into the tensor T. Computer Vision News Tool 13 To have a first view of what the tensor (i.e. face database) looks like, we will use the Slice3() function, simply by typing in the Matlab command: This will bring the following GUI: