EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

The differences in essential facial areas for impressions between humans and deep learning models: An eye‐tracking and explainable AI approach

Takanori Sano iD, Jun Shi, Hideaki Kawabata

British Journal of Psychology · 2026-05 · 117 (2) · 503-527

DOI: 10.1111/bjop.12744 ↗

Kurz gefasst

Fachartikel in der Zeitschrift British Journal of Psychology, online erschienen am 25.10.2024. Verfasst von Takanori Sano und 2 weiteren Autor:innen. Beteiligte Einrichtung: Graduate School of Human Relations, Keio University, Minato-ku, Tokyo, Japan. Hauptthemen laut MEDLINE: Deep Learning, Facial Recognition, Artificial Intelligence, Social Perception. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 66 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: JST SPRING, The Keio University Doctorate Student Grant-in-Aid Program from Ushioda Memorial Fund (Graduate school recommendation), MEXT/JSPS KAKENHI.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study explored the facial impressions of attractiveness, dominance and sexual dimorphism using experimental and computational methods. In Study 1, we generated face images with manipulated morphological features using geometric morphometrics. In Study 2, we conducted eye tracking and impression evaluation experiments using these images to examine how facial features influence impression evaluations and explored differences based on the sex of the face images and participants. In Study 3, we employed deep learning methods, specifically using gradient‐weighted class activation mapping (Grad‐CAM), an explainable artificial intelligence (AI) technique, to extract important features for each impression using the face images and impression evaluation results from Studies 1 and 2. The findings revealed that eye‐tracking and deep learning use different features as cues. In the eye‐tracking experiments, attention was focused on features such as the eyes, nose and mouth, whereas the deep learning analysis highlighted broader features, including eyebrows and superciliary arches. The computational approach using explainable AI suggests that the determinants of facial impressions can be extracted independently of visual attention.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningFacial RecognitionArtificial IntelligenceSocial PerceptionFaceHumansFemaleMaleAdultYoung AdultEye-Tracking Technology

Stichwörter der Autor:innen

deep learningexplainable AIeye‐trackingfacial impressionsgeometric morphometrics

Autor:innen und Institutionen

  • Graduate School of Human Relations, Keio University, Minato-ku, Tokyo, Japan.
  • Faculty of Letters, Keio University, Minato-ku, Tokyo, Japan.

ORCID-Profile: Takanori Sano

Förderung

  • JST SPRING
  • The Keio University Doctorate Student Grant-in-Aid Program from Ushioda Memorial Fund (Graduate school recommendation)
  • MEXT/JSPS KAKENHI
  • Ministry of Education, Culture, Sports, Science and Technology

Zitieren

APA

Sano, T., Shi, J., & Kawabata, H. (2026). The differences in essential facial areas for impressions between humans and deep learning models: An eye‐tracking and explainable AI approach. British Journal of Psychology, 117(2), 503-527. https://doi.org/10.1111/bjop.12744

Vancouver

Sano T, Shi J, Kawabata H. The differences in essential facial areas for impressions between humans and deep learning models: An eye‐tracking and explainable AI approach. British Journal of Psychology. 2026;117(2):503-527. doi:10.1111/bjop.12744

BibTeX
@article{Sano2026,
  author = {Sano, Takanori and Shi, Jun and Kawabata, Hideaki},
  title = {The differences in essential facial areas for impressions between humans and deep learning models: An eye‐tracking and explainable AI approach},
  journal = {British Journal of Psychology},
  year = {2026},
  volume = {117},
  number = {2},
  pages = {503--527},
  doi = {10.1111/bjop.12744}
}
Kostenlos starten →