AI Industry

Facial Recognition in Identity Verification

Facial recognition is a major class of biometric technology which is increasing its market share while expanding its applications across multiple industries.

Facial recognition is a major class of biometric technology which is increasing its market share while expanding its applications across multiple industries. According to a recent report published by Market Research Future (MRFR), the global facial recognition market is predicted to grow at a Compound Annual Growth Rate (CAGR) of 19.68 percent and reach US$8.93 billion by the end of 2022. Facial recognition integrates multiple technologies including Cloud Technology, Internet of Things, Big Data and Automation.

Facial recognition is now mainly used in identity verification to enable accurate comparison and recognition of an individual from a huge database through their facial characteristics. There is no doubt that facial recognition systems have greatly improved the accuracy and efficiency of security verification processes while reducing labour costs, as the technology excels at detecting small but unique facial features that would be overlooked by human eyes.

Facial recognition in payment

Although fingerprints are currently widely accepted in mobile payment systems, the face is the biometric feature of the future in this area. In 2017 Ant Financial cooperated with Face++ to establish the “Smile-to-Pay” system at fast food restaurant KFC in Hangzhou, China. A 3D camera scans customers’ faces as they smile for their chicken. This is a totally automated payment process with no KFC employees required.

An Alibaba employee demonstrates the ‘Smile to Pay’ facial recognition automatic payment system

Facial recognition in customs clearance

In Oct 2018 Hikvision launched a facial verification system which is now integrated into the customs clearance system on the new Hong Kong-Zhuhai-Macao Bridge. The system’s smart facial recognition camera can quickly and accurately recognize a driver’s identity and help staff collect information. With the help of a 180° visual detector, the system can also scan for suspicious activities such as smuggling in seconds, compared to the three minutes required for a human search.

Hikivision camera detector on Hong Kong-Zhuhai-Macao Bridge

Facial recognition in airports

In late 2018 Delta Air Lines partnered with US Customs and Border Protection (CBP), Hartsfield-Jackson Atlanta International Airport (ATL), and the Transportation Security Administration (TSA) to establish a biometric airport terminal system in Atlanta. Facial recognition is used in self-check-in, baggage drop, onboarding, and arrivals. Globally, a number of airports are introducing biometric systems for verifying arriving and departing passengers — with only a passport, fingerprint and face scan required to pass through an automatic security door.

Self check-in machines at Maynard H. Jackson International in Atlanta

Facial recognition systems can deliver convenience, but also bring privacy concerns which have slowed the tech’s wider application. When facial features are recorded there is the possibility this data could be hacked, leaked, or used maliciously to threaten privacy, identity or personal property.

Other concerns regard the robustness of the technology itself. Key facial features could be altered or camouflaged by criminals, some people (such as twins) may possess very similar biometric characteristics, and facial characteristics can change due to age, disease or surgery, etc. It remains to be seen how these issues will impact accuracy when scaled and how facial recognition researchers will resolve them. This is one of the biggest challenges in the technology’s development.

Facial recognition will be increasingly combined with other technologies like big data, automation systems, robots, etc, as humans are taken out of the loop. Trials are already underway in staffless hotels, grocery stores and restaurants worldwide. Smart farms have adopted facial recognition systems for recognizing and recording livestock health and activity data.

If privacy concerns can be effectively mitigated or eliminated and accuracy improved and stabilized, it seems likely that facial recognition systems will become the new standard for identity verification across a growing range of applications.

Author: Shiyuan Jiang | Editor: Michael Sarazen

1 comment on “Facial Recognition in Identity Verification

  1. Pingback: Facial Recognition in Identity Verification – Tech Check News

Leave a Reply

Your email address will not be published. Required fields are marked *

%d bloggers like this: