How Facial Recognition Works

Posted Nov 8, 2024 by Kason Yoo

Nowadays, it’s pretty uncommon for people to type in a password to open their phone. Instead, all they have to do is just look at the screen, and boom, they’re in. This efficient feature is called facial recognition, and everyone’s heard of it, yet not many know how it actually works.

Facial recognition is a way to identify and confirm a person’s identity by using their face, and is used almost everywhere, such as in phones, airports, and law enforcement. It is a form of biometric software. Wait -- what is biometric software? Biometric software is technology that uses a person’s unique physical characteristics to authenticate or identify them. Other examples include fingerprint recognition, voice recognition, and eye retina recognition.

The way facial recognition works is first a photo of the person's face must be presented to the software, and the more angles of the face, the better. Most facial recognition technology relies on 2D images rather than 3D images since it can more conveniently match a 2D image with public photos or those in a database. Next, the software must scan and analyze the face, reading the geometry and plotting facial landmarks to create a map of the specific facial features known as a facial signature. Things the software looks for include distance, depth, and shape.

For more information, watch this video: https://www.youtube.com/watch?v=YX8BzK_LU0E

If you’ve ever seen a picture of a face with a bunch of points on it like a connect the dots, that’s what a facial signature is (see picture below). Key features include the distance between the eyes, the distance from the forehead to the chin, and the shape of the cheekbones.

The next step is to transform the analog information (the face) into digital information (data). The numerical code that’s created is called a faceprint, and just like a fingerprint, they’re unique to each person. If the faceprint presented to the software is a match to the faceprint in its database, then boom. You’re in.

However, while facial recognition is super efficient, it also raises important privacy concerns and can be biased against certain groups of people. As this technology becomes more widely accepted, it’s important to consider the impact on privacy and procedure/policy to ensure fair and unbiased use.

Sources:

https://us.norton.com/blog/iot/how-facial-recognition-software-works

https://www.kaspersky.com/resource-center/definitions/what-is-facial-recognition

Page updated

Report abuse