Related Experiment Video
Updated: Feb 20, 2026
07:34
Perceptual and Category Processing of the Uncanny Valley Hypothesis' Dimension of Human Likeness: Some Methodological Issues
Published on: June 3, 2013
18.0K
Too good to be true: Synthetic AI faces are more average than real faces and super-recognizers know it
James D Dunn1, David White1, Clare A M Sutherland2,3
1School of Psychology, UNSW Sydney, Sydney, New South Wales, Australia.
British Journal of Psychology (London, England : 1953)
|February 18, 2026
Summary
Super-recognizers, individuals with exceptional face recognition skills, are better at distinguishing artificial intelligence (AI) faces from real ones. Their ability is linked to recognizing the
Area of Science:
- Cognitive Psychology
- Artificial Intelligence
- Computer Vision
Background:
- The advancement of artificial intelligence (AI) has led to the creation of highly realistic synthetic faces.
- There is a growing need to differentiate between AI-generated and real human faces due to potential misuse.
- Individual differences in human face recognition abilities may influence the capacity to detect synthetic faces.
Purpose of the Study:
- To investigate whether individual differences in human face recognition ability predict the accuracy in discriminating AI-generated faces from real human faces.
- To explore the relationship between sensitivity to 'hyper-average' facial features and the ability to detect AI faces.
- To examine the role of face-space representation in AI face detection.
Main Methods:
- Comparison of face discrimination abilities between super-recognizers (high face recognition ability) and control groups (typical and high-performing).
- Analysis of the correlation between human face recognition scores and AI face discrimination performance.
- Utilizing deep neural networks to analyze the distribution of AI and real faces in a computational face-space.
Main Results:
- Super-recognizers demonstrated significantly higher accuracy in discriminating AI faces compared to both typical and higher-performing control groups.
- A positive association was found between human face recognition ability and AI face discrimination performance.
- Sensitivity to the 'hyper-average' appearance of AI faces correlated with discrimination ability, particularly for super-recognizers.
Conclusions:
- Exceptional human face recognition abilities enhance the detection of artificial intelligence-generated faces.
- The perception of 'hyper-averageness' serves as a key cue for identifying synthetic faces, especially for individuals with superior face processing skills.
- This study establishes a mechanistic link between evolved face expertise and AI face detection, challenging previous assumptions about human face-space structure.
Related Concept Videos
01:21Association Areas of the Cortex
9.7K
Association areas are regions of the cerebral cortex that do not have a specific sensory or motor function. Instead, they integrate and interpret information from various sources to enable higher cognitive processes such as memory, learning, and decision-making. Some key association areas include the following:
Prefrontal Association Area: This area is located in the frontal lobe and is involved in planning, decision-making, and moderating social behavior. It connects with primary motor areas,...
Prefrontal Association Area: This area is located in the frontal lobe and is involved in planning, decision-making, and moderating social behavior. It connects with primary motor areas,...
9.7K
01:24Prosopagnosia
886
Prosopagnosia, also known as face blindness, is the inability to recognize faces. In severe cases, individuals with prosopagnosia may not recognize close family members, including parents and spouses, by their faces. For instance, someone with prosopagnosia might walk past their child in a crowd, only realizing their mistake upon noticing their child's distinctive backpack or favorite jacket. Prosopagnosia specifically impairs facial recognition, while the recognition of other objects or...
886
01:24Facial Feedback Hypothesis
694
Charles Darwin proposed that facial expressions are an evolutionary adaptation for communication. He argued that these expressions are not influenced by culture but are universal across species. For example, a snarling expression with exposed teeth signals a threat in many animals, including humans. Darwin also suggested that displaying an emotion can intensify the feeling. Smiling, for example, could enhance one's sense of happiness. This idea laid the foundation for understanding the role...
694
01:13Nonconscious Mimicry
5.1K
Nonconscious mimicry occurs when individuals alter their mannerisms to match the behaviors and expressions of those nearby, without intention.
5.1K
02:16Stereotype Content Model
15.5K
The Stereotype Content Model (SCM) was first proposed by Susan Fiske and her colleagues (Fiske, Cuddy, Glick & Xu, 2002; see also Fiske, 2012 and Fiske, 2017). The SCM specifies that when someone encounters a new group, they will stereotype them based on two metrics: warmth—or that group’s perceived intent, and how likely they are to provide help or inflict harm—and competence—or their ability to carry out that objective. Depending on the warmth-competence...
15.5K
01:16Non-equilibrium in the Cell
5.5K
An important concept in studying metabolism and energy is that of chemical equilibrium. Most chemical reactions are reversible. They can proceed in both directions, releasing energy into their environment in one direction, and absorbing it from the environment in the other direction. The same is true for the chemical reactions involved in cell metabolism, such as the breaking down and building up of proteins into and from individual amino acids, respectively. Reactants within a closed system...
5.5K