Dataset Information

Unsupervised inference approach to facial attractiveness.

ABSTRACT: The perception of facial attractiveness is a complex phenomenon which depends on how the observer perceives not only individual facial features, but also their mutual influence and interplay. In the machine learning community, this problem is typically tackled as a problem of regression of the subject-averaged rating assigned to natural faces. However, it has been conjectured that this approach does not capture the complexity of the phenomenon. It has recently been shown that different human subjects can navigate the face-space and "sculpt" their preferred modification of a reference facial portrait. Here we present an unsupervised inference study of the set of sculpted facial vectors in such experiments. We first infer minimal, interpretable and accurate probabilistic models (through Maximum Entropy and artificial neural networks) of the preferred facial variations, that encode the inter-subject variance. The application of such generative models to the supervised classification of the gender of the subject that sculpted the face reveals that it may be predicted with astonishingly high accuracy. We observe that the classification accuracy improves by increasing the order of the non-linear effective interaction. This suggests that the cognitive mechanisms related to facial discrimination in the brain do not involve the positions of single facial landmarks only, but mainly the mutual influence of couples, and even triplets and quadruplets of landmarks. Furthermore, the high prediction accuracy of the subjects' gender suggests that much relevant information regarding the subjects may influence (and be elicited from) their facial preference criteria, in agreement with the multiple motive theory of attractiveness proposed in previous works.

SUBMITTER: Ibanez-Berganza M

PROVIDER: S-EPMC7602690 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Unsupervised inference approach to facial attractiveness.

Ibanez-Berganza Miguel M Amico Ambra A Lancia Gian Luca GL Maggiore Federico F Monechi Bernardo B Loreto Vittorio V

PeerJ 20201028

The perception of facial attractiveness is a complex phenomenon which depends on how the observer perceives not only individual facial features, but also their mutual influence and interplay. In the machine learning community, this problem is typically tackled as a problem of regression of the subject-averaged rating assigned to natural faces. However, it has been conjectured that this approach does not capture the complexity of the phenomenon. It has recently been shown that different human sub ...[more]

PMID: 33194411

Similar Datasets

Project description:ImportanceAesthetic proportions of the lips and their effect on facial attractiveness are poorly defined. Established guidelines would aid practitioners in achieving optimal aesthetic outcomes during cosmetic augmentation.ObjectiveTo assess the most attractive lip dimensions of white women based on attractiveness ranking of surface area, ratio of upper to lower lip, and dimensions of the lip surface area relative to the lower third of the face.Design, setting, and participantsIn phase 1 of this study, synthetic morph frontal digital images of the faces of 20 white women ages 18 to 25 years old were used to generate 5 varied lip surface areas for each face. These 100 faces were cardinally ranked by attractiveness through our developed conventional and internet-based focus groups by 150 participants. A summed ranking score of each face was plotted to quantify the most attractive surface area. In phase 2 of the study, 4 variants for each face were created with 15 of the most attractive images manipulating upper to lower lip ratios while maintaining the most attractive surface area from phase 1. A total of 60 faces were created, and each ratio was ranked by attractiveness by 428 participants (internet-based focus groups). In phase 3, the surface area from the most attractive faces was used to determine the total lip surface area relative to the lower facial third. Data were collected from March 1 to November 31, 2010, and analyzed from June 1 to October 31, 2016.Main outcomes and measuresMost attractive lip surface area, ratio of upper to lower lip, and dimension of the lips relative to the lower facial third.ResultsIn phase 1, all 100 faces were cardinally ranked by 150 individuals (internet-based focus groups [n = 130] and raters from conventional focus groups [conventional raters] [n = 20]). In phase 2, all 60 faces were cardinally ranked by 428 participants (internet-based focus groups [n = 408] and conventional raters [n = 20]). The surface area that corresponded to the range of 2.0 to 2.5 × 104 pixels represented the highest summed rank, generating a pool of 14 images. This surface area was determined to be the most attractive and corresponded to a 53.5% increase in surface area from the original image. With the highest mean and highest proportions of most attractive rankings, the 1:2 ratio was deemed most attractive. Conversely, the ratio of 2:1 was deemed least attractive, having the lowest mean at 1.61 and the highest proportion of ranks within 1 with 310 votes (72.3%).Conclusions and relevanceUsing a robust sample size, this study found that the most attractive lip surface area represents a 53.5% increase from baseline, an upper to lower lip ratio of 1:2, and a surface area equal to 9.6% of the lower third of the face. Lip dimensions and ratios derived in this study may provide guidelines in improving overall facial aesthetics and have clinical relevance to the field of facial plastic surgery.Level of evidenceNA.

Dataset Information

Unsupervised inference approach to facial attractiveness.

Publications

Unsupervised inference approach to facial attractiveness.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets