How is the individuality of a face recognized?

Area of Science:

• Cognitive Psychology
• Neuroscience
• Human Perception

Background:

• Current models of face recognition, including norm-based and exemplar-based approaches, fail to explain rapid, accurate identification in crowds.
• These models rely on comparing observed faces to an average or stored set of faces, which is insufficient for distinguishing individuals.
• The limitations highlight a need for a new model that accounts for the efficiency and accuracy of human face recognition.

Purpose of the Study:

• To propose a modified norm-based model for face recognition that utilizes unusual facial features.
• To explain how the brain constructs a multidimensional 'face-space' based on normal variations of features.
• To demonstrate how identifying rare features enables unique face identification and explains perceptual phenomena like caricatures and the 'own-race effect'.

Main Methods:

• Constructing a multidimensional 'face-space' using normal variations of facial features, excluding those in the extreme 5th and 95th percentiles.
• Defining unusual features as those outside the typical range (below 5th or above 95th percentile).
• Proposing face identification as an 'exception-reporting' process based on the number of unusual features.

Main Results:

• A face with 100 independently variable features, having an average of 10 unusual features, can be rendered unique (1 in 10^13).
• The presence of just 10 rare features is sufficient for a face to be statistically unique within the global population.
• This model explains how exaggerating or attenuating unusual features (caricatures/anti-caricatures) preserves identity.

Conclusions:

• Face recognition is primarily driven by the detection of statistically rare or unusual features, not by comparison to averages.
• The model accounts for the 'own-race effect,' where individuals of a foreign race appear similar due to shared features outside the observer's 'usual range'.
• The evolutionary advantage of recognizing 'usual' features, linked to mate selection, likely underpins this efficient perceptual mechanism.