Trends in Tech: Discovering Facial Recognition Technology

While biometric technology has been around for years, facial recognition technology was once upon a time a concept kept solely for movies about spies, secret government agencies, and top-secret vaults. However, this year saw a significant increase in the use of facial recognition technology in mobile devices and people all over the world are now unlocking their phone with their faces alone. If you thought unlocking your phone with your fingerprint was cool, now you can unlock your phone just by looking at it. Let’s look into facial recognition technology a bit more.

How Does Facial Recognition Work?

Facial recognition technology is defined by the ability to identify a person or verify the identity of a person using only facial details. The facial details are captured, analysed, and compared to patterns in the database and involves 3 steps as outlined below.

  1. Detection: This is an imperative step in the facial recognition process as it distinguishes and pinpoints human faces in both images and videos.
  2. Capture: This process transforms the analogue information of a face into a set of digital information based on the person’s facial features.
  3. Match: This process authenticates whether two faces belong to the same person.

The Facial Recognition Software Race

As with any technological advancement, there are several companies who are vying for the number one spot in facial recognition software race. Here are some of the top scorers:

  • Developed in 2014 by researchers at the Chinese University of Hong Kong, the GaussianFace algorithm achieved facial identification scores of 98.52% in comparison with the 97.53% achieved by humans. Coincidentally, those numbers are also very close to the payout ratios of the best Australian online pokies!

  • Developed by Facebook, the DeepFace programme was announced in 2014 which could determine, with an accuracy rate of 97.25%, whether two photographed faces belonged to the same person.
  • In 2015, Google announced FaceNet which provided unrivalled results: 100% accuracy in the reference test and 95% on the YouTube Faces database.

Putting a Price on Facial Recognition

In a study published in June 2019, it was estimated that by 2024 the global facial recognition market would generate over $7 billion worth of revenue. The two biggest drivers in this exponential growth over the next 5 years can be attributed to surveillance in the public sector and other applications across a diverse collection of market segments. The study noted that the top facial recognition vendors include:

  • Accenture
  • Aware
  • BioID
  • Certibio
  • Fujitsu
  • Fulcrum Biometrics
  • Gemalto
  • HYPR
  • Idemia
  • Leidos
  • M2SYS
  • NEC
  • Nuance
  • Phonexia
  • Smilepass

The Facial Recognition Phenomenon in China

China is the latest country to get in on the facial recognition software trend and is currently being used everywhere from banks to the police force. The police on the outskirts of Beijing have begun using facial recognition sunglasses and a countrywide video surveillance network is currently being set up. Over 200 million cameras are currently being used with that number expected to rise to 626 million by 2020. News has also been released that the country has started making use of a facial recognition payment system.