newsfilter.io
Other

Facial recognition technology will change the way we live

  • 2018 is identified as a landmark year for face recognition, marking the technology's transition from niche development to mainstream adoption.
  • The shift is driven by advancements in machine learning, which enable the extraction of accurate, unique facial prints from photographs.
  • The iPhone's integration of built-in face recognition is cited as the primary catalyst for immediate consumer adoption, with millions of devices expected in circulation by the following year.
  • Beyond smartphone unlocking, the technology is projected to become the standard for accessing banking applications and will begin to feel "normal" to users.
  • In Tel Aviv, engineers have developed a system that translates facial features into text data by measuring distances between eyes, lips, and noses.
  • This software achieves 99% accuracy in controlled environments and can identify a specific individual from millions of records in under one second.
  • The technology allows for covert identification, enabling subjects to be tracked without their knowledge or consent.
  • Commercial applications currently include:
    • Retail stores tracking shopping habits to target in-store advertisements.
    • Medical centers and churches monitoring attendance.
    • A UK school implementing the technology to monitor teacher break times, placing staff on a "watch list."
  • China is described as a global leader in facial recognition deployment, utilizing a government database of 700 million people (half its population).
  • Chinese applications extend beyond security to include:
    • Locating terrorists and individuals with outstanding arrest warrants.
    • Facilitating payments at fast-food restaurants and entry to theme parks without tickets.
    • Attempting to identify individuals stealing public toilet paper.
  • Researchers indicate that facial features can be used to infer sexuality with significant accuracy when analyzed across large datasets (e.g., 10,000 images).
  • The capability to isolate high-probability matches (e.g., "which of these 10 people are most likely to be gay") poses risks for bad actors and authoritarian regimes.
  • The expansion of the technology threatens to fundamentally alter concepts of privacy, fairness, and trust due to the scale of image storage and analysis.
  • Most citizen interactions with face recognition in 2018 will be non-consensual, occurring through security services and retail profiling.
  • Proponents argue that the security and business intelligence benefits outweigh privacy concerns.
  • The proposed solution for managing this new power dynamic is the implementation of regulations to suppress malicious uses while permitting beneficial applications.
  • 2018 is framed as a critical period for defining and establishing these regulatory frameworks.