New AI predicts age by analyzing hands

A new artificial intelligence (AI) facial recognition model developed by Haut.AI, a company that specializes in AI for skin care, reportedly demonstrated the ability to accurately predict a person’s age via hand images.

According to the company, the research demonstrated that AI models trained on hand images achieved accuracy comparable to that of facial recognition technology — and with an average error of 4.1 and 4.7 years in predicting chronological age.

To ensure the representation of diverse skin tones, the researchers trained the AI model on an Indian population dataset, a focus intended to promote fairer AI solutions and to mitigate biases common with conventional systems.

The Haut.AI team employed a type of AI model composed of a convolutional neural network (CNN) to predict an individual’s chronological age according to visible skin features. Using images from 1,454 Indian women between the ages of 20 and 80 from different geographic cohorts with diverse ethnic skin tones, the team randomly split the images into two groups wherein 70% were used for training the CNNs and 30% were used for validation.

The team discovered that the CNN validation demonstrated mean absolute errors of 4.1 and 4.7 years for making predictions about chronological age.

“Collectively, for AI estimates of CA, CNNs based solely on hand images are a viable alternative and comparable to CNNs based on facial images,” the researchers explained.

An article detailing the study, “Predicting human chronological age via AI analysis of dorsal hand versus facial images: A study in a cohort of Indian females,” appears in the journal Experimental Dermatology.

UNLIMITED FREE ACCESS TO THE WORLD’S BEST IDEAS

SIGN UP TO SEE MORE

This is embarrasing…

An error occurred while processing the form. Please try again in a few minutes.

Customize Your GlobalSpec Experience

This is embarrasing…

An error occurred while processing the form. Please try again in a few minutes.