MATICLASS: Understanding Face Match

ben-sweet-2LowviVHZ-E-unsplash
Share on facebook
Share on twitter
Share on linkedin

Hello everyone, and welcome back to a new MatiClass – hope you’re as excited as I am! This time we’re going to focus on one of the most useful types of checks an Identity verification platform like Mati can perform. Face match, also known as Face Detection or Face Recognition. So without any further ado, let’s begin our lesson:

What is Face Match

In the identity verification industry, Face Match belongs to the family of the biometric checks. The name pretty much explains what it is. But basically: computer technology that compares an image containing a face to one or more facial images. This establishes whether the faces likely belong to the same person, therefore verifying whether they are considered a match

This is a prominent feature for our clients since it’s super useful to ensure that the person in question is the same as in the documents submitted. It’s also one of the best ways to prevent fraudsters from accessing your platform. To discover what else can help here, I strongly suggest that you read this article.

How Face Match Works In 5 Easy Steps:

  1. Every time you’re asking your users to upload a picture or a Liveness video of themselves, a face match check can be run. An algorithm extracts a set of selfies from the images submitted and then determines which ones contain a clear (non-blurry) face.
  1. The selfie or selfies are then compared to the identification that the user previously submitted. Let’s say their passport, national ID, or driving license; basically, any official document containing a picture of their face. To do so, an algorithm first extracts the facial visual from the document submitted and then compares it to the selfie using a neural network trained on millions of face images.
  1. The algorithm’s output is a set of intermediate similarity scores that are then concentrated to produce a final similarity score.
  1. This similarity score is compared to a vast array of scores generated by millions of verification processes. From there, your Identity verification solution determines if the pair of faces belong to the same person or doesn’t. 

 Red: Imposter Pairs

Green: Authentic Pairs

  1. Lastly, you receive the final result in your back end and dashboard. -Did the user successfully passed the Face match or did he fail?

NB: at Mati, we believe you should be able to set your own threshold to define if someone passed the Face Match check or didn’t. Indeed, depending on your industry and the type of products you’re selling, you may want to implement a strict Face Match process. This process should have a minimum similarity score of 70% or a looser process. Note that our recommendation at Mati is to always implement a high Face Match score rule since your chances of banning fraudsters from your service will be higher.

Why Is Face Match A Necessary Step To Onboard Users?

We use Face Match technology for a wide range of purposes. One of the most important being  KYC (Know Your Customer) processes during account creation or onboarding. If you’re running a digital service such as a payment app,  gig economy platforms, a neobank app, or a lending service, then you are fully aware of KYC compliance and what a headache these can be when done manually. 

Let me just say that during these COVID times, technology is truly our ally. Another advantage as well is that the check is entirely automated. If you ever bootstrapped a product before or created complex manual processes, you probably see where I’m headed. Face Match is an excellent way to automate time-consuming videocalls. It’s also a perfect way to increase your own level of confidence in these processes. After all, “To err is human”- validating someone’s identity via a messy skype call, with a bad internet connection and awful video quality is really not the best way to make sure the person you have in front of your screen matches the one in the document. 

For this reason, it is important to understand how it works and how it could transform your own industry. Fraudsters can easily exploit today’s technological advances; therefore, liveness detection is vital for modern biometric-based recognition solutions.

Reach out to one of our Face Match experts to learn more.

Compártelo con tus amigos

Share on facebook
Share on twitter
Share on linkedin

Deja un comentario:

Your email address will not be published.

Suscríbete a nuestro blog