Detecting artificial facial images using facial landmarks

A system including at least one device for classifying facial images, the device comprising:

• a processor configured to execute instructions stored in a memory of the device;
• a classifier trained to classify facial images as real facial images or as artificial facial images;
• a reader to read a provided facial image associated with a subject;
• and
• a scanner to obtain a captured facial image of the subject,
• wherein the processor is further configured to:
• retrieve a first plurality of facial landmarks of the provided facial image associated with the subject;
• retrieve a corresponding second plurality of facial landmarks of the captured facial image of the subject;
• extract at least one feature vector based on differences between at least some of the first plurality of facial landmarks and corresponding facial landmarks of the second plurality of facial landmarks, wherein at least some values of the at least one feature vector represent angular values of facial landmarks;
• and
• supply the at least one feature vector to the classifier to classify the provided facial image as a real facial image of the subject or as an artificial facial image,
• wherein each facial landmark in the first and second plurality of facial landmarks has an associated facial landmark in the same plurality of facial landmarks, and
• wherein the processor is further configured to determine, for each facial landmark, a landmark vector between the facial landmark and the associated facial landmark.

A method for classifying facial images, comprising:

• retrieving a first plurality of facial landmarks of a provided facial image associated with a subject;
• determining a corresponding second plurality of facial landmarks of a captured facial image of the subject;
• extracting at least one feature vector based on differences between at least some of the first plurality of facial landmarks and corresponding facial landmarks of the second plurality of facial landmarks, wherein at least some values of the at least one feature vector represent angular values of facial landmarks;
• and
• supplying the at least one feature vector to a classifier to classify the provided facial image as a real facial image of the subject or as an artificial facial image,
• wherein each facial landmark in the first plurality of facial landmarks and the second plurality of facial landmarks has an associated facial landmark in the same plurality of facial landmarks, the method further comprising determining, for each facial landmark, a landmark vector between the facial landmark and the associated facial landmark.