Skip to content
Last Dictionary
Plane
Vision, Speech & PerceptionSociety, Ethics & Economy

face detection

nounAlso: face detector, Viola–Jones detector, Haar cascade, facial landmark detection

In plain wordsFinding where the faces are in a photo or video, without working out whose faces they are: the step that lets a camera focus on people or blur them out.

  1. The task of locating every human face in an image or video frame, usually as a bounding boxbounding box, n.: The rectangle a vision system draws around an object it has found, stored as a handful of numbers giving the rectangle's position and size. plus a few landmarks such as the eyes, nose tip and mouth corners. It is a special case of object detectionobject detection, n.: Finding every object of interest in a picture or video frame, saying what each one is and drawing a box around it, for example all the cars, cyclists and pedestrians in a street scene. and the first stage of most face-analysis pipelines, which go on to alignment and then recognition, attribute or expression analysis. The Viola–Jones detector (2001) made it practical in real time: rectangle-contrast (Haar-like) features were read in constant time from a precomputed 'integral image', AdaBoostAdaBoost, prop. n.: One of the first practical boosting methods: it trains simple models in turn, giving extra weight to examples the earlier models got wrong, then lets all the models vote, weighted by accuracy. combined the most useful ones into classifiers, and chaining these let most windows be rejected after a few cheap tests. It was built into digital cameras and the OpenCV library. Deep detectors (MTCNN, 2016; RetinaFace, 2019; small mobile models) handle extreme poses, occlusion and tiny faces, and are benchmarked on datasets such as WIDER FACE. Uses include camera autofocus and exposure, photo organization, privacy blurring and video calls; error rates that differ by skin tone, age or lighting are a recognized fairness concern.

    Run face detection first and blur every box before the street-level images leave the vehicle.

Origin

Automatic face analysis dates to the 1960s–70s (Woody Bledsoe's work; Takeo Kanade's 1973 thesis). Neural-network detectors by Henry Rowley, Shumeet Baluja and Kanade (1998) were followed by Paul Viola and Michael Jones's boosted-cascade detector, presented at CVPR 2001, which made face detection ubiquitous.

Usage note

Detection only finds faces; facial recognitionfacial recognition, n.: Technology that identifies or checks who someone is by comparing a picture of their face with stored photos, used by phones, passport gates and police databases. determines whose face it is. Most biometric regulation, including the EU AI ActEU AI Act, prop. n.: The European Union's law on artificial intelligence, the first broad AI law anywhere: it bans some uses, sets strict checks for risky ones and puts duties on the makers of the biggest AI models.'s rules on remote biometric identificationremote biometric identification, n.: Using AI to recognize people at a distance — usually by matching faces in camera footage against a database — without their taking part. The EU strictly limits live police use of it in public places., concerns recognition and categorization rather than detection alone.

Related

See also
  • object detectionobject detection, n.: Finding every object of interest in a picture or video frame, saying what each one is and drawing a box around it, for example all the cars, cyclists and pedestrians in a street scene.
  • facial recognitionfacial recognition, n.: Technology that identifies or checks who someone is by comparing a picture of their face with stored photos, used by phones, passport gates and police databases.
  • keypoint detectionkeypoint detection, n.: Finding specific, recognizable points in an image, such as the corners of a building, a person's elbows and knees, or the tip of a nose.
  • AdaBoostAdaBoost, prop. n.: One of the first practical boosting methods: it trains simple models in turn, giving extra weight to examples the earlier models got wrong, then lets all the models vote, weighted by accuracy.
  • biometric categorisationbiometric categorisation, n.: Using AI to sort people into groups — by age, sex, ethnicity and the like — from their faces, voices or other body features. The EU bans using it to guess sensitive traits such as religion or sexuality.
  • algorithmic biasalgorithmic bias, n.: Unfairness in software that judges or serves people: it consistently produces worse results for some groups, such as women, older people or ethnic minorities, often because it learned from skewed or incomplete records.
Contrast with
  • facial recognitionfacial recognition, n.: Technology that identifies or checks who someone is by comparing a picture of their face with stored photos, used by phones, passport gates and police databases.

Sources

  1. Viola, P. & Jones, M. (2001). Rapid Object Detection Using a Boosted Cascade of Simple Features. CVPR 2001.
  2. Viola, P. & Jones, M. J. (2004). Robust Real-Time Face Detection. International Journal of Computer Vision 57(2), 137–154.

Mentioned in

object detectionobject detection, n.: Finding every object of interest in a picture or video frame, saying what each one is and drawing a box around it, for example all the cars, cyclists and pedestrians in a street scene., AdaBoostAdaBoost, prop. n.: One of the first practical boosting methods: it trains simple models in turn, giving extra weight to examples the earlier models got wrong, then lets all the models vote, weighted by accuracy., keypoint detectionkeypoint detection, n.: Finding specific, recognizable points in an image, such as the corners of a building, a person's elbows and knees, or the tip of a nose.