The Estonia-based business, Haut.AI – which specialises in highly personalised skincare and works with top beauty businesses such as Beiersdorf and Ulta Beauty – said that using hands instead of faces helps to “address potential biases that can arise from facial recognition systems due to factors like ethnicity and facial features.”
It highlighted that the research emphasised the importance of using diverse datasets in AI development to ensure unbiased, more inclusive solutions and that the technology offers an alternative for situations where facial images are unavailable or less preferred.
Haut AI’s CEO Anastasia Georgievskaya shared: “Our research demonstrates that your age can be determined just as accurately from a picture of your hands as from your face.”
She added that “This not only opens doors for new applications of AI technology, but also has the potential to mitigate biases often associated with conventional systems.”
The study showed that AI models trained on hand images had comparable accuracy to those using facial images, with an average error of 4.1 and 4.7 years in predicting chronological age.
Significant for ethnic skin
Georgievskaya said that the research is particularly significant for ethnic skin, as it was trained using the Indian population dataset. It also “represents the first AI model for age prediction specifically designed with a diverse dataset that includes a wide range of skin tones.”
The AI company added that the study goes beyond just accuracy of age prediction, stating: “By analysing how specific features on hands and faces influence the model’s predictions, the research contributes to a better understanding of the ageing process.”
The researchers discovered that the areas of skin around the eyes, nose, mouth, and forehead were important for facial age prediction by AI. These areas often show wrinkles, sagging, and other signs of ageing. While on the hands, features like wrinkles, knuckles, and bone prominence were significant for age prediction.
Haut AI said that anti-ageing interventions that address these features will make you look younger to neural networks and, most likely, to humans, too.