A New Hypothesis on Facial Beauty Perception

Abstract

In this article, a new hypothesis on facial beauty perception is proposed: the weighted average of two facial geometric features is more attractive than the inferior one between them. Extensive evidences support the new hypothesis. We collected 390 well-known beautiful face images (e.g., Miss Universe, movie stars, and super models) as well as 409 common face images from multiple sources. Dozens of volunteers rated the face images according to their attractiveness. Statistical regression models are trained on this database. Under the empirical risk principle, the hypothesis is tested on 318,801 pairs of images and receives consistently supportive results. A corollary of the hypothesis is attractive facial geometric features construct a convex set. This corollary derives a convex hull based face beautification method, which guarantees attractiveness and minimizes the before-after difference. Experimental results show its superiority to state-of-the-art geometric based face beautification methods. Moreover, the mainstream hypotheses on facial beauty perception (e.g., the averageness, symmetry, and golden ratio hypotheses) are proved to be compatible with the proposed hypothesis.

Analysis of android application based on images in cellular phone

“…Face symmetry alone cannot use to decide the attractiveness rate. Mostly, face that considered as beautiful have non-perfect symmetry 8 or average face. 5 How hot are you have 9 about 36% more accurate than nFace and Golden Ratio Face Rater which only 15%.…”

Data-Driven Facial Beauty Analysis: Prediction, Retrieval and Manipulation

“…12.7e, g keep the above details and are easier to identify. It may be argued that the attractiveness (Chen et al 2014). c Model-based geometry (MG) beautification (Leyvand et al 2008).…”

“…Chen et al (2014) proved that convex combinations of beautiful face shapes are guaranteed to be beautiful, i.e., if B ¼ fxjf ðxÞ ! y 0 g, where x is the shape feature, then f ð P x i 2B h i x i Þ !…”

“…The above is called exemplar-based manipulation. It is designed for facial geometry manipulation, with the weighted average hypothesis proved in Chen et al (2014) as the groundwork. This approach is not suitable for appearance features, because convex combinations of appearance features will lose the identity information of the query face.…”

“…We built a database that contains 1,609 high-quality female face images, which is an extension of that used in our previous work (Chen et al 2014). The images include well-known beautiful faces collected from the web (e.g., Miss Universe, Miss World, movie stars, and super models) and facial images from several existing databases (e.g., the XM2VTS database Messer et al 1999, FERET database Phillips et al 2000, and the Shanghai database Zhang et al 2011).…”