Techniques for robust anti-spoofing in biometrics using polarization cues for NIR and visible wavelength band

A non-transitory computer-readable medium having computer-executable instructions stored thereon that, in response to execution by one or more processors of a computing system, cause the computing system to perform actions for training a machine learning model to verify a biometric measurement of a physical feature of a subject, the actions comprising:

• collecting, by the computing system, a set of training examples, wherein collecting each training example includes receiving polarization information, two-dimensional image information, and an indication of whether the subject is spoofed in the training example;
• and
• training, by the computing system, the machine learning model to verify biometric measurements using the set of training examples;
• wherein training the machine learning model to verify biometric measurements using the set of training examples includes, for each training example of the set of training examples, using the indication of whether the subject is spoofed in the training example as an expected output of the machine learning model for the training example.

A non-transitory computer-readable medium having computer-executable instructions stored thereon that, in response to execution by one or more processors of a computing system, cause the computing system to perform actions for training a machine learning model to validate a biometric measurement, the actions comprising:

• collecting, by the computing system, a set of training examples, wherein collecting each training example includes receiving polarization information, two-dimensional image information, and an indication of whether a subject is spoofed in the training example;
• and
• training, by the computing system, the machine learning model to verify whether biometric measurements are spoofed using the set of training examples;
• wherein training the machine learning model to verify whether biometric measurements are spoofed using the set of training examples includes, for each training example of the set of training examples, using the indication of whether the subject is spoofed in the training example as an expected output of the machine learning model for the training example.