Bryan McAllister’s personal reference notebook
September 12, 2026Short & citedBlog
Civic Viewpoint.
A quiet encyclopedia with a personal desk
How It Works • Artificial Intelligence

Facial recognition

Facial recognition is a biometric technology that analyzes facial features in images or video to identify or verify a person’s identity.

Updated September 12, 20261 min readCited sources

Facial recognition is a biometric technology that identifies or verifies people by analyzing facial features. It's used in devices and security systems, but raises privacy and bias concerns.

What it is

Facial recognition systems typically detect a face in an image or video, align it, extract distinctive features such as the distance between facial landmarks, and convert those into a numerical template. That template is then compared to a stored database to verify or identify a person.

Key details

The technology can be used for one-to-one verification (e.g., unlocking a phone) or one-to-many identification (e.g., matching a face against a watchlist). Common uses include device unlocking, access control, identity verification, and some law enforcement and commercial applications. Accuracy varies by system, lighting, image quality, and demographic group; errors and false matches remain possible even as the technology improves.

Why it matters

Facial recognition raises privacy and surveillance concerns because facial data can be collected and stored at scale, and errors may amplify bias or lead to false positives. In Canada, biometric data such as facial information is generally subject to privacy law, and legality depends on the organization, purpose, consent, and compliance with applicable privacy requirements.

In short

  • Facial recognition analyzes facial features to identify or verify people.
  • Accuracy varies by system and demographic group; errors can occur.
  • Raises privacy, surveillance, and bias concerns.
  • Canadian use is governed by privacy law and consent rules.
Canadian angle

Facial recognition is relevant to Canadian readers because Canadian public bodies and private organizations may use it for security and identification, but its use is subject to Canadian privacy and consent rules, making governance and oversight central issues.

Quick questions

How does face recognition work?
It detects a face, measures facial landmarks, converts them into a numerical template, and compares that template with stored data to verify or identify a person.
How accurate is face recognition?
Accuracy varies widely by system and conditions. Performance is improving, but errors, false positives, and demographic disparities can still occur.
Is face recognition legal in Canada?
It can be legal, but its use is constrained by privacy law and oversight. Whether a specific use is lawful depends on the organization, purpose, collection method, and compliance with Canadian privacy requirements.

Sources

  1. Wikipediahttps://en.wikipedia.org/wiki/Facial_recognition_system
    Supports: Four-step process: face detection, alignment, feature extraction, and matching; general definition and applications.
  2. Signicathttps://www.signicat.com/blog/face-recognition
    Supports: Biometric authentication description; four-step workflow; facial templates and use cases.
  3. Center for Strategic and International Studieshttps://www.csis.org/analysis/how-does-facial-recognition-work
    Supports: Comparison of face images using computer-generated filters and deep learning.
  4. The Conversationhttps://theconversation.com/facial-recognition-is-increasingly-common-but-how-does-it-work-61354
    Supports: Common uses, privacy concerns, and discussion of error rates and accuracy limits.
  5. Oktahttps://www.okta.com/identity-101/how-does-facial-recognition-work/
    Supports: Matching process, database comparison, and note that accuracy depends on input data.
  6. Courserahttps://www.coursera.org/articles/what-is-facial-recognition
    Supports: General definition and process of detection, analysis, and recognition.