Face Match – Remove the Guesswork Out of ‘Identity Fraud Detection’

Our facial recognition program can determine if two photos are actually of the same person.

Get a live photo of someone’s face, make sure it’s real, and then compare that photo to the photo in their ID or passport to make sure they match. Further, our adaptable AI face engine may be deployed On-Premise with your in-house applications to automate manual face identification operations, boosting efficiency and slashing the potential for fraud.

Let’s Define the Obvious

“Facial Match” is the process of determining or establishing the identity of a person based on their facial features. Identifying people in photographs, videos, or even in real time is possible with the help of facial recognition systems.
Did you know that certain features of a face never change? Like, for example, the eyballs of a person remain the same since birth. The eyeball is the only organism which does not grow from birth. It is fully grown when you are born. When you look at a baby’s face, you see mostly iris and little white. As the baby grows, you get to see more and more of the eyeball. There are 16 such markers in a face that never change irrespective of ageing. Our deep learning algorithm compares these markers to give an accuracy match.

How is it Used in Banking, Financial Services and Insurance Industries

Typically, when you are getting a service registration or delivery from any of the BFSI industries, the KYC – Know Your Customer verification process kicks in.This process requires that the customer submit their identification, something like the

  • Aadhar Card
  • Passport
  • Voter ID
  • PAN Card

Anything that establishes the identity of the customer and also their address. Of course, the identity establishment is what Face Match concerns itself with.

The problem really presents itself when the ID proof and the identity become difficult, as at times, the ID proof is a document that is very old and the person could be visibly different from the presented ID document.

ATM fraud prevention in India is being handled by banks using a facial recognition technology currently in use. In addition to that, it is utilized for the reporting of voters who have already cast their ballots, the verification of passports and visas, driving licenses, and so on.

We use 16 different facial markers to establish that the id and person are the same. The 16 face markers include

  • Upper and Lower Eye Lids
  • The Eyebrow
  • Nose Bridge
  • Upper and Lower Lips
  • Chin
  • Jaw Bone

These are aspects of a face that does not vary with age (for adults) and remain a consistent part of their facial features throughout their lifespan.

How Robust is Facial Recognition?

This technology is being used all over the world, and it has brought about many positive outcomes. China has already implemented several of them.

Us, we have implemented these systems with the following clients

We have addressed several use cases and are continuously evolving our solution for better market fitment as the demand increases.

Solution Key Features

Vision Technology Enabled Search / Face detection

Within a matter of seconds, our algorithm is able to map out the contours of two images of the same person and determine the face match and recommend the degree of accuracy of the presented documents against each other.

A person’s facial structure can be mapped, identified, or verified with the help of a technology using facial recognition. It is possible to generate a one-of-a-kind numerical code using face recognition technology. This code is referred to as a faceprint.

These faceprints are going to be stored in a database that deals with face recognition. If you upload a photo to the database, it will search for any faceprints that are a match and display them to you.

Quality Management of Images

How do you determine the quality of an image? The computer vision looks at an image and provides detailed information on the image used. The quality is assessed with relation to the positions of the eye and mouth, adequate landmarks, blur, reflection, clarity, exposure, orientation (pitch, yaw, roll), and adequate landmarks.

Reading from a very poor source

Though the system does what it says, i.e. verify that two faces belong to the same person by determining the likelihood that they belong to the same person, the kicker is that the images can be in any format, badly presented which is a common problem with many of identification documents, as the quality deteriorates as time passes and is largely dependent on the person maintaining the document.

Our algorithms are capable of identifying the minute details of the facemarkers and present functionally useful information for further decision making.

Liveness Verification for Authentication

If there’s a way, rest assured that someone will find it to fool the system. An image can be used to unlock access points and our system understands the difference between an image and live person, through its advanced AI based learning algorithm.

We ensure Data Anonymity

Images are anonymously transformed into a string of alphanumeric characters (binary) that cannot be deciphered backwards. The hash storage is one of the strongest encryption methods prevalent and yet to be broken technologies and is available as a de facto with our solution.

Do You Need a BFSI or KYC Related Solution For your app or business?

You can utilize our technology exactly how you envision it, regardless of whether you are just starting out or are already leagues ahead of the competition and want to optimize the experience you provide for your customers. in a variety of methods and for a wide range of applications.

The solution is already being used by some of the top Insurance brands in the market and is proven beyond doubt.

Book A Demo

Because it will allow for more efficient management, establishments such as educational institutions, academic institutions, and even medical care facilities are planning to install facial recognition technology on their premises. Complex technology, such as that utilized in facial technology, is increasingly making its way into the robotics industry.
Book a Demo