Forming the Ability of Recognition of Faces in AI-Generated Photos
https://doi.org/10.18384/3033-6414-2026-2-92-104
Abstract
Aim. To develop a training strategy that enhances the ability to distinguish between real and AI-generated photos.
Methodology. The study involved 201 respondents aged from 14 to 66 years, students and specialists of various profiles (m-45, f-156, Mage=23.66, SD=8.65). The study was conducted as a formative experiment with initial and final measurements. The experimental group participated in training based on a synthesis of feedback strategies and direct instruction. Quantitative data were processed using the Mann-Whitney U-test and Wilcoxon test in Statsoft STATISTICA 10.0.
Results. At the initial measurement, the experimental and control groups were equivalent. Statistically significant differences were found between the initial and final measurements in the experimental group. There were also statistically significant differences in the control group, though to a lesser extent than in the experimental group.
Research implications. The training strategy for recognizing AI-generated photos, based on combining feedback and direct instruction strategies, indeed improves participants’ detection accuracy. The results may be useful for further research in the field of perception and interaction with AI.
Supporting agencies: This study was supported by the Ministry of Science and Higher Education of the Russian Federation within the frame-work of a State assignment (FEWZ-2024-0052)
About the Authors
I. V. Vasileva
University of Tyumen; Tyumen Institute for Advanced Training of Employees of the Ministry of Internal Affairs of Russia
Russian Federation
Russian Federation
Inna V. Vasileva (Tyumen) – Dr. Sci. (Psychology), Assoc. Prof., Head of the Department, Department of Psychology and Defectology, School of Education ; Prof., Department of Philosophy, Foreign Languages, Humanitarian Training
A. N. Asafova
Independent researcher
Russian Federation
Russian Federation
Anastasia N. Asafova (Tyumen) – Independent researcher
D. S. Grischenko
Independent researcher
Russian Federation
Russian Federation
Darya S. Grischenko (Tyumen) – Independent researcher
References
1. Katsyri, J., de Gelder, B. & Takala, T. (2019). Virtual Faces Evoke Only a Weak Uncanny Valley Effect: An Empirical Investigation With Controlled Virtual Face Images. In: Perception, 48 (10), 968–991. DOI: 10.1177/0301006619869134.
2. Natale, D. A., Simonetti, M. E. & La Rocca, S. (2023). Uncanny Valley Effect: A Qualitative Synthesis of Empirical Research to Assess the Suitability of Using Virtual Faces in Psychological Research. In: Computers in Human Behavior Reports, 10 (3). URL: https://www.sciencedirect.com (accessed: 10.10.2025). DOI: 10.1016/j.chbr.2023.100288.
3. Miller, E. J., Foo, Y. Z. & Mewton, P. et al. (2023). How Do People Respond to Computer-generated Versus Human Faces? A Systematic Review and Meta-analyses. In: Computers in Human Behavior Reports, 10. URL: https://www.sciencedirect.com (accessed: 10.10.2025). DOI: 10.1016/j.chbr.2023.100283.
4. Korteling, J. E., van de Boer-Visschedijk, G. C. et al. (2021). Human- Versus Artificial Intelligence. In: Frontiers in Artificial Intelligence, 4. URL: https://www.frontiersin.org (accessed: 10.10.2025). DOI: 10.3389/frai.2021.622364.
5. Köbis, N. C., Doležalová, B. & Soraperra, I. (2021). Fooled Twice: People Cannot Detect Deepfakes but Think They Can. In: iScience, 24 (11). URL: https://www.cell.com/iscience (accessed: 10.10.2025).
6. Rathgeb, C., Tolosana, R. & Vera-Rodriguez, R. et al. (2022). Handbook of Digital Face Manipulation and Detection. From Deepfakes to Morphing attacks. Boston: Springer publ. DOI: 10.1007/978-3-030-87664-7.
7. Connolly, B. (2020). Digital Trust: Social Media Strategies to Increase Trust and Engage Customers. London: Bloomsbury Business publ.
8. Fletcher, J. (2018). Deepfakes, Artificial Intelligence, and Some Kind of Dystopia: The New Faces of Online Post-Fact Performance. In: Theater Journal, 70, 4, 455–471. DOI: 10.1353/tj.2018.0097.
9. Willis, J. & Todorov, A. (2006). First Impression: Making Up Your Mind After a 100-Ms Exposure to a Face. In: Psychological Science, 17 (7), 592–598. DOI: 10.1111/j.1467-9280.2006.01750.x.
10. Nemrodov, D., Niemeier, M. & Mok, J. N. Y. et al. (2016). The Time Course of Individual Face Recognition: A Pattern Analysis of ERP Signals. In: NeuroImage, 132, 469–476. DOI: 10.1016/j.neuroimage.2016.03.006.
11. Miller, E. J., Steward, B. A. & Witkower, Z. et al. (2023). AI Hyperrealism: Why AI Faces Are Perceived as More Real Than Human Ones. In: Psychological Science, 34 (12), 1390–1403. DOI: 10.1177/09567976231207095.
12. Barabanshchikov, V. A. (2009). Perception of Facial Expressions. Moscow: Institute of Psychology of the RAS publ. (in Russ.).
13. Diel, A., Lalgi, T. & Schroeter, I. C. et al. (2024). As Good as Chance: A Systematic Review and Meta-analysis of Human Deepfake Detection Performance Based on 56 Papers. In: Expert Systems with Applications, 252. URL: https://www.sciencedirect.com (accessed: 10.10.2025). DOI: 10.31219/osf.io/cxv4r.
14. Diel, A., Teufel, M. & Bäuerle, A. (2024). Inability to Detect Deepfakes: Deepfake Detection Training Improves Detection Accuracy, but Increases Emotional Distress and Reduces Selfefficacy. In: OSF Preprints. URL: https://osf.io/preprints/osf (accessed: 10.10.2025). DOI: 10.2139/ssrn.5002585.
15. Mader, B., Banks, M. S. & Farid, H. (2017). Identifying Computer-generated Portraits: The Importance of Training and Incentives. In: Perception, 46 (9), 1062–1076. DOI: 10.1177/0301006617713633.
16. Hulzebosch, N., Ibrahimi, S. & Worring, M. (2020). Detecting CNN-generated Facial Images in Real-world Scenarios. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. URL: https://arxiv.org/abs/2005.05632 (accessed: 10.10.2025). DOI: 10.1109/CVPRW50498.2020.00329.
17. Nightingale, S. J. & Farid, H. (2022). Ai-synthesized Faces Are Indistinguishable From Real Faces and More Trustworthy. In: Proceedings of the National Academy of Sciences, 119 (8). URL: https://www.pnas.org (accessed: 10.10.2025). DOI: 10.1073/pnas.2120481119.
18. Mohamed, N. B., Bogdanel, G. & Moreno, H. G. (2025). Is Training Useful to Detect Deepfakes? A Preliminary Study. In: Royal Society Open Science, 12. URL: https://royalsocietypublishing.org (accessed: 10.10.2025). DOI: 10.23919/CISTI58278.2023.10211622.
19. Bray, S. D., Johnson, S. D. & Kleinberg, B. (2023). Testing Human Ability to Detect ‘deepfake’ Images of Human Faces. In: Journal of Cybersecurity, 9 (1). URL: https://academic.oup.com/cybersecurity (accessed: 10.10.2025). DOI: 10.1093/cybsec/tyad011.
20. Somoray, K. & Miller, D. J. (2023). Providing Detection Strategies to Improve Human Detection of Deepfakes: An Experimental Study. In: Computers in Human Behavior, 149 (4). URL: https://www.sciencedirect.com (accessed: 10.10.2025). DOI: 10.1016/j.chb.2023.107917.
21. Yegemberdiyeva, G. & Amirgaliyev, B. (2021). Study Of AI Generated and Real Face Perception. In: Proceedings of the 2021 IEEE International Conference on Smart Information Systems and Technologies (SIST). 2021. URL: https://ieeexplore.ieee.org (accessed: 10.10.2025). DOI: 10.1109/SIST50301.2021.9465908.
22. Vasilyeva, I. V., Asafova, A. N. & Grishchenko, D. S. (2025). Communicative Characteristics of Personality and the Evaluation of Photographs Generated by Artificial Intelligence. In: Theoretical and Experimental Psychology, 18 (1), 27–44. DOI: 10.11621/TEP-25-02 (in Russ.).
23. Devi, S. P. & Mishra, A. K. (2023). The Testing Effect: Looking Through Implicit Theories’ Perspectives. In: Psychological Studies, 68 (1), 92–100. DOI: 10.1007/s12646-022-00710-6
24. Ng, Y.–L. (2023). An Error Management Approach to Perceived Fakeness of Deepfakes: The Moderating Role of Perceived Deepfake Targeted Politicians’ Personality Characteristics. In: Current Psychology, 42 (1), 25658–25669. DOI: 10.1007/s12144-022-03621-x.
25. Schetinger, V., Oliveira, M. M. & da Silva, R. et al. (2015). Humans Are Easily Fooled by Digital Images. In: Computers & Graphics, 68. URL: https://arxiv.org/abs/1509.05301 (accessed: 10.10.2025). DOI: 10.48550/arXiv.1509.05301.
Review
Views:
56
JATS XML
Email this article 