Synthetic and natural face identity processing share common mechanisms

Recent developments in generative AI offer the means to create synthetic identities, or deepfakes, at scale. As deepfake faces and voices become indistinguishable from real ones, they are considered as promising alternatives for research and development to enhance fairness and protect humans' r...

Full description

Saved in:
Bibliographic Details
Main Authors: Kim Uittenhove, Hatef Otroshi Shahreza, Sébastien Marcel, Meike Ramon
Format: Article
Language:English
Published: Elsevier 2025-03-01
Series:Computers in Human Behavior Reports
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2451958824001969
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850044472638832640
author Kim Uittenhove
Hatef Otroshi Shahreza
Sébastien Marcel
Meike Ramon
author_facet Kim Uittenhove
Hatef Otroshi Shahreza
Sébastien Marcel
Meike Ramon
author_sort Kim Uittenhove
collection DOAJ
description Recent developments in generative AI offer the means to create synthetic identities, or deepfakes, at scale. As deepfake faces and voices become indistinguishable from real ones, they are considered as promising alternatives for research and development to enhance fairness and protect humans' rights to privacy. Notwithstanding these efforts and intentions, a basic question remains unanswered: Are natural faces and facial deepfakes perceived and remembered in the same way? Using images created via professional photography on the one hand, and a state-of-the-art generative model on the other, we investigated the most studied process of face cognition: perceptual matching and discrimination of facial identity. Our results demonstrate that identity discrimination of natural and synthetic faces is governed by the same underlying perceptual mechanisms: objective stimulus similarity and observers’ ability level. These findings provide empirical support both for the societal risks associated with deepfakes, while also underscoring the utility of synthetic identities for research and development.
format Article
id doaj-art-513867cbcf9a434ba7b8e9d2ee97e37a
institution DOAJ
issn 2451-9588
language English
publishDate 2025-03-01
publisher Elsevier
record_format Article
series Computers in Human Behavior Reports
spelling doaj-art-513867cbcf9a434ba7b8e9d2ee97e37a2025-08-20T02:54:58ZengElsevierComputers in Human Behavior Reports2451-95882025-03-011710056310.1016/j.chbr.2024.100563Synthetic and natural face identity processing share common mechanismsKim Uittenhove0Hatef Otroshi Shahreza1Sébastien Marcel2Meike Ramon3Center for Learning Science, EPFL, Lausanne, Switzerland; Applied Face Cognition Lab, Institute of Psychology, University of Lausanne, SwitzerlandIdiap Research Institute, Martigny, Switzerland; School of Engineering, EPFL, Lausanne, SwitzerlandIdiap Research Institute, Martigny, Switzerland; School of Criminal Justice, University of Lausanne, SwitzerlandApplied Face Cognition Lab, Institute of Psychology, University of Lausanne, Switzerland; AIR – Association for Independent Research, Zurich, Switzerland; Corresponding author. Applied Face Cognition Lab, Institute of Psychology, University of Lausanne, Switzerland.Recent developments in generative AI offer the means to create synthetic identities, or deepfakes, at scale. As deepfake faces and voices become indistinguishable from real ones, they are considered as promising alternatives for research and development to enhance fairness and protect humans' rights to privacy. Notwithstanding these efforts and intentions, a basic question remains unanswered: Are natural faces and facial deepfakes perceived and remembered in the same way? Using images created via professional photography on the one hand, and a state-of-the-art generative model on the other, we investigated the most studied process of face cognition: perceptual matching and discrimination of facial identity. Our results demonstrate that identity discrimination of natural and synthetic faces is governed by the same underlying perceptual mechanisms: objective stimulus similarity and observers’ ability level. These findings provide empirical support both for the societal risks associated with deepfakes, while also underscoring the utility of synthetic identities for research and development.http://www.sciencedirect.com/science/article/pii/S2451958824001969Face identity processingNatural and synthetic facesDeepfakesStimulus similarityHuman abilityIndividual differences
spellingShingle Kim Uittenhove
Hatef Otroshi Shahreza
Sébastien Marcel
Meike Ramon
Synthetic and natural face identity processing share common mechanisms
Computers in Human Behavior Reports
Face identity processing
Natural and synthetic faces
Deepfakes
Stimulus similarity
Human ability
Individual differences
title Synthetic and natural face identity processing share common mechanisms
title_full Synthetic and natural face identity processing share common mechanisms
title_fullStr Synthetic and natural face identity processing share common mechanisms
title_full_unstemmed Synthetic and natural face identity processing share common mechanisms
title_short Synthetic and natural face identity processing share common mechanisms
title_sort synthetic and natural face identity processing share common mechanisms
topic Face identity processing
Natural and synthetic faces
Deepfakes
Stimulus similarity
Human ability
Individual differences
url http://www.sciencedirect.com/science/article/pii/S2451958824001969
work_keys_str_mv AT kimuittenhove syntheticandnaturalfaceidentityprocessingsharecommonmechanisms
AT hatefotroshishahreza syntheticandnaturalfaceidentityprocessingsharecommonmechanisms
AT sebastienmarcel syntheticandnaturalfaceidentityprocessingsharecommonmechanisms
AT meikeramon syntheticandnaturalfaceidentityprocessingsharecommonmechanisms