A Study Shows AI Technique Can Differentiate Smiles Based On Gender

A new technique has been developed by a team of researchers rooted on AI (artificial intelligence) for charting a smile’s dynamics that can differentiate between women and men automatically. The research team from the University of Bradford in the United Kingdom plotted 49 markers on the face, particularly around the mouth, eyes, and down the nose.

They utilized these to evaluate how the face alters as we smile resulted from the movements of the underlying muscle—comprising both variation in distances between the diverse spots and the smile’s “flow”: how far, how fast, and how much the diverse face points shifted as the smile was produced. Then they validated whether there were visible variations between women and men—and discovered that there were, with smiles of women being more extensive.

Hassan Ugail, Professor from the University of Bradford, said, “Anecdotally, females are considered to be more communicative in how they smile, and our study has confirmed this. Females certainly have wider smiles, extending their lip and mouth area far more compared to males.” For the research, the team developed an algorithm making use of their study and validated it against the video footage of about 109 individuals as they smiled. The study is published in the International Journal of Computer Graphics: The Visual Computer.

The computer was capable of appropriately finding out the gender in 86% of cases and the researchers deem the accuracy can be enhanced easily. Ugail said, “As this system evaluates the face’s fundamental muscle movement during a smile, we consider these dynamics will stay the same, though the exterior physical features vary, subsequent to surgery, for instance.”

“This sort of facial recognition can be a future biometric, as it is not reliant on one trait, but on a dynamic that is exclusive to a person and will be extremely hard to alter or mimic,” he added.