Biometric facial recognition technology is software that scans a person's facial characteristics and then compares that against a scourge image, determining a percentage match. Most commonly, facial recognition is used for authentication and verification of a person's identity through an app.
Biometric recognition software will be able to identify individuals from their facial features with 90%+ accuracy. This technology is commonly used in mobile phone facial recognition, such as Apple's Face ID. However, it is becoming increasingly common for this kind of software to be used within the Know Your Customer KYC processes that regulated businesses perform on potential clients.
Facial recognition software advancements have enabled streamlined biometric identification checks, such as AML and KYC aspects of the onboarding process. With an increase in online fraud and digital hacks, it has become more critical than ever to verify the identities of individuals and save the business and the UK economy from the adverse effects of money laundering and identity fraud.
Each human body has unique integrated patterns, including ear shape and eye colour. Recognition software measures each of these patterns in the same way that a fingerprint is measured, and so can a face. These features and patterns are called biometric modalities.
Biometric facial recognition works by taking these biometric modalities in a video or a picture of an individual and comparing it to another, like a passport or a driver's licence picture, that biometric data is then used to compare against photos and videos.
It does this by reconstructing images on a three-dimensional level and matching it to a two-dimensional image. This technology uses five modules to perform this task.
It does this by locating a human face within an image, subtracting any background or complex backgrounds and creating two components, labelled "face" and "non-face". Then the face is digitally aligned using facial landmarks such as eyes, ears and nose. These are like fingerprints on the face; comparison software then matches each of these features to a reference image, comparing each part of the facial structure until a conclusion of a match is made or not.
Because facial recognition is so accurate without the downside of being invasive on any level, it has become one of the primary forms of biometric identity verification, and the technology has now become so widely applicable.
As so many people have access to smartphones and thus access to a camera, the ability to deploy this software as an app and web app means that any business can utilise this technology. Unlike a fingerprint, DNA and iris scanners need specialised hardware and the individual to be present. Other forms of biometric modalities are.
During the initial Know Your Customer (KYC) and Anti-Money Laundering (AML) checks that new clients undergo, identifying and verifying their identity can sometimes be a lengthy and costly. By automating this process, you can not only reduce time and cost but also improve the accuracy of your results.
Validient has introduced a liveness check step in their verification software; this utilises a biometric facial recognition app, comparing the result of that check against a provided image from a driving licence or passport; we then determine if there is a match and notify you of our results.
With identity fraud and money laundering becoming more sophisticated, traditional detection and prevention methods are no longer adequate. The adoption of automated and biometric verification software is becoming the only way to tackle this fundamental problem.
You can learn more about verifying a client's identity here.
Biometric facial recognition can serve a positive purpose when identifying and verifying potential clients; there can be instances where this technology can be missed. Specifically with misidentification, this usually means facial recognition software mistaking someone for someone else. However, there can be a malicious use for this.
The act of face spoofing is where a fraudulent user or a criminal will use pictures and pre-recorded videos to fool facial recognition software into believing they are the genuine person. This is an attempt to gain access to accounts, like banks and even access to mobile phones and other forms of financial crimes. You can read more about Identity Theft here.
We at Validient, however, combat this by asking users to upload photographic IDs to cross-check and also ask users to record a video with unique one-time words or numbers; this prevents criminals from being able to fool the system.
The other issue is, what if a client doesn't have a photographic ID? We have discussed this scenario here if you want to find out more.
A biometric facial recognition app is a fantastic opportunity to speed up your Know Your Customer process and allow you to be up-to-date with AML compliance and KYC compliance. With technology being very accurate and reliable, the benefits it brings to a business are clear. Reduce time, cost and the risk of working with new clients. Validient offer a robust full Know Your Customer product that identifies and verifies clients complete with face recognition technology.